barman-2.1/ 0000775 0001750 0001750 00000000000 13033432341 013206 5 ustar gcalacoci gcalacoci barman-2.1/barman/ 0000775 0001750 0001750 00000000000 13033432341 014446 5 ustar gcalacoci gcalacoci barman-2.1/barman/exceptions.py 0000664 0001750 0001750 00000015521 13015546105 017211 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
class BarmanException(Exception):
"""
The base class of all other barman exceptions
"""
class ConfigurationException(BarmanException):
"""
Base exception for all the Configuration errors
"""
class CommandException(BarmanException):
"""
Base exception for all the errors related to
the execution of a Command.
"""
class CompressionException(BarmanException):
"""
Base exception for all the errors related to
the execution of a compression action.
"""
class PostgresException(BarmanException):
"""
Base exception for all the errors related to PostgreSQL.
"""
class BackupException(BarmanException):
"""
Base exception for all the errors related to the execution of a backup.
"""
class WALFileException(BarmanException):
"""
Base exception for all the errors related to WAL files.
"""
class HookScriptException(BarmanException):
"""
Base exception for all the errors related to Hook Script execution.
"""
class LockFileException(BarmanException):
"""
Base exception for lock related errors
"""
class DuplicateWalFile(WALFileException):
"""
A duplicate WAL file has been found
"""
class MatchingDuplicateWalFile(DuplicateWalFile):
"""
A duplicate WAL file has been found, but it's identical to the one we
already have.
"""
class SshCommandException(CommandException):
"""
Error parsing ssh_command parameter
"""
class UnknownBackupIdException(BackupException):
"""
The searched backup_id doesn't exists
"""
class BackupInfoBadInitialisation(BackupException):
"""
Exception for a bad initialization error
"""
class CommandFailedException(CommandException):
"""
Exception representing a failed command
"""
class CommandMaxRetryExceeded(CommandFailedException):
"""
A command with retry_times > 0 has exceeded the number of available retry
"""
def __init__(self, exc):
"""
:param Exception exc: the last exception raised by the command
"""
self.exc = exc
super(CommandMaxRetryExceeded, self).__init__(*exc.args)
class RsyncListFilesFailure(CommandException):
"""
Failure parsing the output of a "rsync --list-only" command
"""
class DataTransferFailure(CommandException):
"""
Used to pass failure details from a data transfer Command
"""
@classmethod
def from_command_error(cls, cmd, e, msg):
"""
This method build a DataTransferFailure exception and report the
provided message to the user (both console and log file) along with
the output of the failed command.
:param str cmd: The command that failed the transfer
:param CommandFailedException e: The exception we are handling
:param str msg: a descriptive message on what we are trying to do
:return DataTransferFailure: will contain the message provided in msg
"""
details = msg
details += "\n%s error:\n" % cmd
details += e.args[0]['out']
details += e.args[0]['err']
return cls(details)
class CompressionIncompatibility(CompressionException):
"""
Exception for compression incompatibility
"""
class FsOperationFailed(CommandException):
"""
Exception which represents a failed execution of a command on FS
"""
class LockFileBusy(LockFileException):
"""
Raised when a lock file is not free
"""
class LockFilePermissionDenied(LockFileException):
"""
Raised when a lock file is not accessible
"""
class LockFileParsingError(LockFileException):
"""
Raised when the content of the lockfile is unexpected
"""
class ConninfoException(ConfigurationException):
"""
Error for missing or failed parsing of the conninfo parameter (DSN)
"""
class PostgresConnectionError(PostgresException):
"""
Error connecting to the PostgreSQL server
"""
def __str__(self):
# Returns the first line
if self.args and self.args[0]:
return str(self.args[0]).splitlines()[0].strip()
else:
return ''
class PostgresAppNameError(PostgresConnectionError):
"""
Error setting application name with PostgreSQL server
"""
class PostgresSuperuserRequired(PostgresException):
"""
Superuser access is required
"""
class PostgresIsInRecovery(PostgresException):
"""
PostgreSQL is in recovery, so no write operations are allowed
"""
class PostgresUnsupportedFeature(PostgresException):
"""
Unsupported feature
"""
class PostgresDuplicateReplicationSlot(PostgresException):
"""
The creation of a physical replication slot failed because
the slot already exists
"""
class PostgresReplicationSlotsFull(PostgresException):
"""
The creation of a physical replication slot failed because
the all the replication slots have been taken
"""
class PostgresReplicationSlotInUse(PostgresException):
"""
The drop of a physical replication slot failed because
the replication slots is in use
"""
class PostgresInvalidReplicationSlot(PostgresException):
"""
Exception representing a failure during the deletion of a non
existent replication slot
"""
class TimeoutError(CommandException):
"""
A timeout occurred.
"""
class ArchiverFailure(WALFileException):
"""
Exception representing a failure during the execution
of the archive process
"""
class BadXlogSegmentName(WALFileException):
"""
Exception for a bad xlog name
"""
class BadHistoryFileContents(WALFileException):
"""
Exception for a corrupted history file
"""
class AbortedRetryHookScript(HookScriptException):
"""
Exception for handling abort of retry hook scripts
"""
def __init__(self, hook):
"""
Initialise the exception with hook script info
"""
self.hook = hook
def __str__(self):
"""
String representation
"""
return ("Abort '%s_%s' retry hook script (%s, exit code: %d)" % (
self.hook.phase, self.hook.name,
self.hook.script, self.hook.exit_status))
barman-2.1/barman/copy_controller.py 0000664 0001750 0001750 00000070005 13026732452 020247 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
Copy controller module
A copy controller will handle the copy between a series of files and directory,
and their final destination.
"""
import collections
import logging
import os.path
import re
import shutil
import tempfile
from functools import partial
import dateutil.parser
import dateutil.tz
from barman.command_wrappers import RsyncPgData
from barman.exceptions import CommandFailedException, RsyncListFilesFailure
_logger = logging.getLogger(__name__)
class _RsyncCopyItem(object):
"""
Internal data object that contains the information about one of the items
that have to be copied during a RsyncCopyController run.
"""
def __init__(self, label, src, dst,
exclude=None,
exclude_and_protect=None,
include=None,
is_directory=False,
bwlimit=None,
reuse=None,
item_class=None,
optional=False):
"""
The "label" parameter is meant to be used for error messages
and logging.
If "src" or "dst" content begin with a ':' character, it is a remote
path. Only local paths are supported in "reuse" argument.
If "reuse" parameter is provided and is not None, it is used to
implement the incremental copy. This only works if "is_directory" is
True
:param str label: a symbolic name for this item
:param str src: source directory.
:param str dst: destination directory.
:param list[str] exclude: list of patterns to be excluded from the
copy. The destination will be deleted if present.
:param list[str] exclude_and_protect: list of patterns to be excluded
from the copy. The destination will be preserved if present.
:param list[str] include: list of patterns to be included in the
copy even if excluded.
:param bool is_directory: Whether the item points to a directory.
:param bwlimit: bandwidth limit to be enforced. (KiB)
:param str|None reuse: the reference path for incremental mode.
:param string|None item_class: If specified carries a meta information
about what the object to be copied is.
:param bool optional: Whether a failure copying this object should be
treated as a fatal failure. This only works if "is_directory" is
False
"""
self.label = label
self.src = src
self.dst = dst
self.exclude = exclude
self.exclude_and_protect = exclude_and_protect
self.include = include
self.is_directory = is_directory
self.bwlimit = bwlimit
self.reuse = reuse
self.item_class = item_class
self.optional = optional
# Attributes that will e filled during the analysis
self.dir_file = None
self.safe_file = None
self.check_file = None
self.exclude_and_protect_file = None
def __str__(self):
# Prepare strings for messages
formatted_class = self.item_class
formatted_name = self.src
if self.src.startswith(':'):
formatted_class = 'remote ' + self.item_class
formatted_name = self.src[1:]
formatted_class += ' directory' if self.is_directory else ' file'
# Log the operation that is being executed
if self.item_class in(RsyncCopyController.PGDATA_CLASS,
RsyncCopyController.PGCONTROL_CLASS):
return "%s: %s" % (
formatted_class, formatted_name)
else:
return "%s '%s': %s" % (
formatted_class, self.label, formatted_name)
class RsyncCopyController(object):
"""
Copy a list of files and directory to their final destination.
"""
# Constants to be used as "item_class" values
PGDATA_CLASS = "PGDATA"
TABLESPACE_CLASS = "tablespace"
PGCONTROL_CLASS = "pg_control"
CONFIG_CLASS = "config"
# This regular expression is used to parse each line of the output
# of a "rsync --list-only" call. This regexp has been tested with any known
# version of upstream rsync that is supported (>= 3.0.4)
LIST_ONLY_RE = re.compile('''
(?x) # Enable verbose mode
^ # start of the line
# capture the mode (es. "-rw-------")
(?P[-\w]+)
\s+
# size is an integer
(?P\d+)
\s+
# The date field can have two different form
(?P
# "2014/06/05 18:00:00" if the sending rsync is compiled
# with HAVE_STRFTIME
[\d/]+\s+[\d:]+
|
# "Thu Jun 5 18:00:00 2014" otherwise
\w+\s+\w+\s+\d+\s+[\d:]+\s+\d+
)
\s+
# all the remaining characters are part of filename
(?P.+)
$ # end of the line
''')
# This regular expression is used to ignore error messages regarding
# vanished files that are not really an error. It is used because
# in some cases rsync reports it with exit code 23 which could also mean
# a fatal error
VANISHED_RE = re.compile('''
(?x) # Enable verbose mode
^ # start of the line
(
# files which vanished before rsync start
rsync:\ link_stat\ ".+"\ failed:\ No\ such\ file\ or\ directory\ \(2\)
|
# files which vanished after rsync start
file\ has\ vanished:\ ".+"
|
# files which have been truncated during transfer
rsync:\ read\ errors\ mapping\ ".+":\ No\ data\ available\ \(61\)
|
# final summary
rsync\ error:\ .* \(code\ 23\)\ at\ main\.c\(\d+\)
\ \[generator=[^\]]+\]
)
$ # end of the line
''')
# This named tuple is used to parse each line of the output
# of a "rsync --list-only" call
_FileItem = collections.namedtuple('FileItem', 'mode size date path')
def __init__(self, path=None, ssh_command=None, ssh_options=None,
network_compression=False,
reuse_backup=None, safe_horizon=None,
exclude=None, retry_times=0, retry_sleep=0):
"""
:param str|None path: the PATH where rsync executable will be searched
:param str|None ssh_command: the ssh executable to be used
to access remote paths
:param list[str]|None ssh_options: list of ssh options to be used
to access remote paths
:param boolean network_compression: whether to use the network
compression
:param str|None reuse_backup: if "link" or "copy" enables
the incremental copy feature
:param datetime.datetime|None safe_horizon: if set, assumes that every
files older than it are save to copy without checksum verification.
:param list[str]|None exclude: list of patterns to be excluded
from the copy
:param int retry_times: The number of times to retry a failed operation
:param int retry_sleep: Sleep time between two retry
"""
super(RsyncCopyController, self).__init__()
self.path = path
self.ssh_command = ssh_command
self.ssh_options = ssh_options
self.network_compression = network_compression
self.reuse_backup = reuse_backup
self.safe_horizon = safe_horizon
self.exclude = exclude
self.retry_times = retry_times
self.retry_sleep = retry_sleep
self.item_list = []
"""List of items to be copied"""
self.rsync_cache = {}
"""A cache of RsyncPgData objects"""
# Attributes used for progress reporting
self.total_steps = None
"""Total number of steps"""
self.current_step = None
"""Current step number"""
def add_directory(self, label, src, dst,
exclude=None,
exclude_and_protect=None,
include=None,
bwlimit=None, reuse=None, item_class=None):
"""
Add a directory that we want to copy.
If "src" or "dst" content begin with a ':' character, it is a remote
path. Only local paths are supported in "reuse" argument.
If "reuse" parameter is provided and is not None, it is used to
implement the incremental copy. This only works if "is_directory" is
True
:param str label: symbolic name to be used for error messages
and logging.
:param str src: source directory.
:param str dst: destination directory.
:param list[str] exclude: list of patterns to be excluded from the
copy. The destination will be deleted if present.
:param list[str] exclude_and_protect: list of patterns to be excluded
from the copy. The destination will be preserved if present.
:param list[str] include: list of patterns to be included in the
copy even if excluded.
:param bwlimit: bandwidth limit to be enforced. (KiB)
:param str|None reuse: the reference path for incremental mode.
:param string item_class: If specified carries a meta information about
what the object to be copied is.
"""
self.item_list.append(
_RsyncCopyItem(
label=label,
src=src,
dst=dst,
is_directory=True,
bwlimit=bwlimit,
reuse=reuse,
item_class=item_class,
optional=False,
exclude=exclude,
exclude_and_protect=exclude_and_protect,
include=include))
def add_file(self, label, src, dst, item_class=None, optional=False):
"""
Add a file that we want to copy
:param str label: symbolic name to be used for error messages
and logging.
:param str src: source directory.
:param str dst: destination directory.
:param bwlimit: bandwidth limit to be enforced. (KiB)
:param string item_class: If specified carries a meta information about
what the object to be copied is.
:param bool optional: Whether a failure copying this object should be
treated as a fatal failure.
"""
self.item_list.append(
_RsyncCopyItem(
label=label,
src=src,
dst=dst,
is_directory=False,
bwlimit=None,
reuse=None,
item_class=item_class,
optional=optional))
def rsync_factory(self, item):
"""
Build the RsyncPgData object required for copying the provided item
:param _RsyncCopyItem item: information about a copy operation
:rtype: RsyncPgData
"""
# If the object already exists, use it
if item in self.rsync_cache:
return self.rsync_cache[item]
# Prepare the command arguments
args = self._reuse_args(item.reuse)
# Merge the global exclude with the one into the item object
if self.exclude and item.exclude:
exclude = self.exclude + item.exclude
else:
exclude = self.exclude or item.exclude
# TODO: remove debug output or use it to progress tracking
# By adding a double '--itemize-changes' option, the rsync
# output will contain the full list of files that have been
# touched, even those that have not changed
args.append('--itemize-changes')
args.append('--itemize-changes')
# Build the rsync object that will execute the copy
rsync = RsyncPgData(
path=self.path,
ssh=self.ssh_command,
ssh_options=self.ssh_options,
args=args,
bwlimit=item.bwlimit,
network_compression=self.network_compression,
exclude=exclude,
exclude_and_protect=item.exclude_and_protect,
include=item.include,
retry_times=self.retry_times,
retry_sleep=self.retry_sleep,
retry_handler=partial(self._retry_handler, item)
)
self.rsync_cache[item] = rsync
return rsync
def copy(self):
"""
Execute the actual copy
"""
# Create a temporary directory to hold the file lists.
temp_dir = tempfile.mkdtemp(suffix='', prefix='barman-')
# The following try block is to make sure the temporary directory
# will be removed on exit.
try:
# Initialize the counters used by progress reporting
self._progress_init()
_logger.info("Copy started (safe before %r)", self.safe_horizon)
# Execute some preliminary steps for each item to be copied
for item in self.item_list:
# The initial preparation is necessary only for directories
if not item.is_directory:
continue
# Analyze the source and destination directory content
self._progress_report("analyze %s" % item)
self._analyze_directory(item, temp_dir)
# Prepare the target directories, removing any unneeded file
self._progress_report(
"create destination directories and delete unknown files "
"for %s" % item)
self._create_dir_and_purge(item)
# Do the actual copy
for item in self.item_list:
# Build the rsync object required for the copy
rsync = self.rsync_factory(item)
# If the item is a directory use the copy method,
# otherwise run a plain rsync
if item.is_directory:
# Log the operation that is being executed
self._progress_report("copy safe files from %s" %
item)
# Copy the safe files
self._copy(rsync,
item.src,
item.dst,
file_list=item.safe_file,
checksum=False)
# Log the operation that is being executed
self._progress_report("copy files with checksum from %s" %
item)
# Copy the files with rsync
self._copy(rsync,
item.src,
item.dst,
file_list=item.check_file,
checksum=True)
else:
# Log the operation that is being executed
self._progress_report("copy %s" % item)
rsync(item.src, item.dst, allowed_retval=(0, 23, 24))
if rsync.ret == 23:
if item.optional:
_logger.warning(
"Ignoring error reading %s", item)
else:
raise CommandFailedException(dict(
ret=rsync.ret, out=rsync.out, err=rsync.err))
finally:
# Clean tmp dir and log, exception management is delegeted to
# the executor class
shutil.rmtree(temp_dir)
_logger.info("Copy finished (safe before %s)", self.safe_horizon)
def _progress_init(self):
"""
Init counters used by progress logging
"""
self.total_steps = 0
for item in self.item_list:
# Directories require 4 steps, files only one
if item.is_directory:
self.total_steps += 4
else:
self.total_steps += 1
self.current_step = 0
def _progress_report(self, msg):
"""
Log a message containing the progress
:param str msg: the message
"""
self.current_step += 1
_logger.info("Copy step %s of %s: %s",
self.current_step, self.total_steps, msg)
def _reuse_args(self, reuse_directory):
"""
If reuse_backup is 'copy' or 'link', build the rsync option to enable
the reuse, otherwise returns an empty list
:param str reuse_directory: the local path with data to be reused
:rtype: list[str]
"""
if self.reuse_backup in ('copy', 'link') and \
reuse_directory is not None:
return ['--%s-dest=%s' % (self.reuse_backup, reuse_directory)]
else:
return []
def _retry_handler(self, item, command, args, kwargs, attempt, exc):
"""
:param _RsyncCopyItem item: The item that is being processed
:param RsyncPgData command: Command object being executed
:param list args: command args
:param dict kwargs: command kwargs
:param int attempt: attempt number (starting from 0)
:param CommandFailedException exc: the exception which caused the
failure
"""
_logger.warn("Failure executing rsync on %s (attempt %s)",
item, attempt)
_logger.warn("Retrying in %s seconds", self.retry_sleep)
def _analyze_directory(self, item, temp_dir):
"""
Analyzes the status of source and destination directories identifying
the files that are safe from the point of view of a PostgreSQL backup.
The safe_horizon value is the timestamp of the beginning of the
older backup involved in copy (as source or destination). Any files
updated after that timestamp, must be checked as they could have been
modified during the backup - and we do not reply WAL files to update
them.
The destination directory must exist.
If the "safe_horizon" parameter is None, we cannot make any
assumptions about what can be considered "safe", so we must check
everything with checksums enabled.
If "ref" parameter is provided and is not None, it is looked up
instead of the "dst" dir. This is useful when we are copying files
using '--link-dest' and '--copy-dest' rsync options.
In this case, both the "dst" and "ref" dir must exist and
the "dst" dir must be empty.
If source or destination path begin with a ':' character,
it is a remote path. Only local paths are supported in "ref" argument.
:param _RsyncCopyItem item: information about a copy operation
:param temp_dir: path to a temporary directory used to store file lists
"""
# Build the rsync object required for the analysis
rsync = self.rsync_factory(item)
# If reference is not set we use dst as reference path
ref = item.reuse
if ref is None:
ref = item.dst
# Make sure the ref path ends with a '/' or rsync will add the
# last path component to all the returned items during listing
if ref[-1] != '/':
ref += '/'
# Build a hash containing all files present on reference directory.
# Directories are not included
try:
ref_hash = dict((
(item.path, item)
for item in self._list_files(rsync, ref)
if item.mode[0] != 'd'))
except (CommandFailedException, RsyncListFilesFailure) as e:
# Here we set ref_hash to None, thus disable the code that marks as
# "safe matching" those destination files with different time or
# size, even if newer than "safe_horizon". As a result, all files
# newer than "safe_horizon" will be checked through checksums.
ref_hash = None
_logger.error(
"Unable to retrieve reference directory file list. "
"Using only source file information to decide which files"
" need to be copied with checksums enabled: %s" % e)
# The 'dir.list' file will contain every directory in the
# source tree
item.dir_file = os.path.join(temp_dir, '%s_dir.list' % item.label)
dir_list = open(item.dir_file, 'w+')
# The 'safe.list' file will contain all files older than
# safe_horizon, as well as files that we know rsync will
# check anyway due to a difference in mtime or size
item.safe_file = os.path.join(temp_dir, '%s_safe.list' % item.label)
safe_list = open(item.safe_file, 'w+')
# The 'check.list' file will contain all files that need
# to be copied with checksum option enabled
item.check_file = os.path.join(temp_dir, '%s_check.list' % item.label)
check_list = open(item.check_file, 'w+')
# The 'protect.list' file will contain a filter rule to protect
# each file present in the source tree. It will be used during
# the first phase to delete all the extra files on destination.
item.exclude_and_protect_file = os.path.join(
temp_dir, '%s_exclude_and_protect.filter' % item.label)
exclude_and_protect_filter = open(item.exclude_and_protect_file,
'w+')
for item in self._list_files(rsync, item.src):
# If item is a directory, we only need to save it in 'dir.list'
if item.mode[0] == 'd':
dir_list.write(item.path + '\n')
continue
# Add every file in the source path to the list of files
# to be protected from deletion ('exclude_and_protect.filter')
exclude_and_protect_filter.write('P ' + item.path + '\n')
exclude_and_protect_filter.write('- ' + item.path + '\n')
# If source item is older than safe_horizon,
# add it to 'safe.list'
if self.safe_horizon and item.date < self.safe_horizon:
safe_list.write(item.path + '\n')
continue
# If ref_hash is None, it means we failed to retrieve the
# destination file list. We assume the only safe way is to
# check every file that is older than safe_horizon
if ref_hash is None:
check_list.write(item.path + '\n')
continue
# If source file differs by time or size from the matching
# destination, rsync will discover the difference in any case.
# It is then safe to skip checksum check here.
dst_item = ref_hash.get(item.path, None)
if (dst_item is None or dst_item.size != item.size or
dst_item.date != item.date):
safe_list.write(item.path + '\n')
continue
# All remaining files must be checked with checksums enabled
check_list.write(item.path + '\n')
# Close all the control files
dir_list.close()
safe_list.close()
check_list.close()
exclude_and_protect_filter.close()
def _create_dir_and_purge(self, item):
"""
Create destination directories and delete any unknown file
:param _RsyncCopyItem item: information about a copy operation
"""
# Build the rsync object required for the analysis
rsync = self.rsync_factory(item)
# Create directories and delete any unknown file
self._rsync_ignore_vanished_files(
rsync,
'--recursive',
'--delete',
'--files-from=%s' % item.dir_file,
'--filter', 'merge %s' % item.exclude_and_protect_file,
item.src, item.dst,
check=True)
def _copy(self, rsync, src, dst, file_list, checksum=False):
"""
The method execute the call to rsync, using as source a
a list of files, and adding the the checksum option if required by the
caller.
:param Rsync rsync: the Rsync object used to retrieve the list of files
inside the directories
for copy purposes
:param str src: source directory
:param str dst: destination directory
:param str file_list: path to the file containing the sources for rsync
:param bool checksum: if checksum argument for rsync is required
"""
# Build the rsync call args
args = ['--files-from=%s' % file_list]
if checksum:
# Add checksum option if needed
args.append('--checksum')
self._rsync_ignore_vanished_files(rsync, src, dst, *args, check=True)
def _list_files(self, rsync, path):
"""
This method recursively retrieves a list of files contained in a
directory, either local or remote (if starts with ':')
:param Rsync rsync: the Rsync object used to retrieve the list
:param str path: the path we want to inspect
:except CommandFailedException: if rsync call fails
:except RsyncListFilesFailure: if rsync output can't be parsed
"""
_logger.debug("list_files: %r", path)
# Use the --no-human-readable option to avoid digit groupings
# in "size" field with rsync >= 3.1.0.
# Ref: http://ftp.samba.org/pub/rsync/src/rsync-3.1.0-NEWS
rsync.get_output('--no-human-readable', '--list-only', '-r', path,
check=True)
for line in rsync.out.splitlines():
line = line.rstrip()
match = self.LIST_ONLY_RE.match(line)
if match:
mode = match.group('mode')
# no exceptions here: the regexp forces 'size' to be an integer
size = int(match.group('size'))
try:
date = dateutil.parser.parse(match.group('date'))
date = date.replace(tzinfo=dateutil.tz.tzlocal())
except (TypeError, ValueError):
# This should not happen, due to the regexp
msg = ("Unable to parse rsync --list-only output line "
"(date): '%s'" % line)
_logger.exception(msg)
raise RsyncListFilesFailure(msg)
path = match.group('path')
yield self._FileItem(mode, size, date, path)
else:
# This is a hard error, as we are unable to parse the output
# of rsync. It can only happen with a modified or unknown
# rsync version (perhaps newer than 3.1?)
msg = ("Unable to parse rsync --list-only output line: "
"'%s'" % line)
_logger.error(msg)
raise RsyncListFilesFailure(msg)
def _rsync_ignore_vanished_files(self, rsync, *args, **kwargs):
"""
Wrap an Rsync.get_output() call and ignore missing args
TODO: when rsync 3.1 will be widespread, replace this
with --ignore-missing-args argument
:param Rsync rsync: the Rsync object used to execute the copy
"""
kwargs['allowed_retval'] = (0, 23, 24)
rsync.get_output(*args, **kwargs)
# If return code is 23 and there is any error which doesn't match
# the VANISHED_RE regexp raise an error
if rsync.ret == 23 and rsync.err is not None:
for line in rsync.err.splitlines():
match = self.VANISHED_RE.match(line.rstrip())
if match:
continue
else:
_logger.error("First rsync error line: %s", line)
raise CommandFailedException(dict(
ret=rsync.ret, out=rsync.out, err=rsync.err))
return rsync.out, rsync.err
barman-2.1/barman/process.py 0000664 0001750 0001750 00000013335 13015546105 016507 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see
import errno
import logging
import os
import signal
import time
from glob import glob
from barman import output
from barman.exceptions import LockFileParsingError
from barman.lockfile import ServerWalReceiveLock
_logger = logging.getLogger(__name__)
class ProcessInfo(object):
"""
Barman process representation
"""
def __init__(self, pid, server_name, task):
"""
This object contains all the information required to identify a
barman process
:param int pid: Process ID
:param string server_name: Name of the server owning the process
:param string task: Task name (receive-wal, archive-wal...)
"""
self.pid = pid
self.server_name = server_name
self.task = task
class ProcessManager(object):
"""
Class for the management of barman processes owned by a server
"""
# Map containing the tasks we want to retrieve (and eventually manage)
TASKS = {
'receive-wal': ServerWalReceiveLock
}
def __init__(self, config):
"""
Build a ProcessManager for the provided server
:param config: configuration of the server owning the process manager
"""
self.config = config
self.process_list = []
# Cycle over the lock files in the lock directory for this server
for path in glob(os.path.join(self.config.barman_lock_directory,
'.%s-*.lock' % self.config.name)):
for task, lock_class in self.TASKS.items():
# Check the lock_name against the lock class
lock = lock_class.build_if_matches(path)
if lock:
try:
# Use the lock to get the owner pid
pid = lock.get_owner_pid()
except LockFileParsingError:
_logger.warning(
"Skipping the %s process for server %s: "
"Error reading the PID from lock file '%s'",
task, self.config.name, path)
break
# If there is a pid save it in the process list
if pid:
self.process_list.append(
ProcessInfo(pid, config.name, task))
# In any case, we found a match, so we must stop iterating
# over the task types and handle the the next path
break
def list(self, task_filter=None):
"""
Returns a list of processes owned by this server
If no filter is provided, all the processes are returned.
:param str task_filter: Type of process we want to retrieve
:return list[ProcessInfo]: List of processes for the server
"""
server_tasks = []
for process in self.process_list:
# Filter the processes if necessary
if task_filter and process.task != task_filter:
continue
server_tasks.append(process)
return server_tasks
def kill(self, process_info, retries=10):
"""
Kill a process
Returns True if killed successfully False otherwise
:param ProcessInfo process_info: representation of the process
we want to kill
:param int retries: number of times the method will check
if the process is still alive
:rtype: bool
"""
# Try to kill the process
try:
_logger.debug("Sending SIGINT to PID %s", process_info.pid)
os.kill(process_info.pid, signal.SIGINT)
_logger.debug("os.kill call succeeded")
except OSError as e:
_logger.debug("os.kill call failed: %s", e)
# The process doesn't exists. It has probably just terminated.
if e.errno == errno.ESRCH:
return True
# Something unexpected has happened
output.error("%s", e)
return False
# Check if the process have been killed. the fastest (and maybe safest)
# way is to send a kill with 0 as signal.
# If the method returns an OSError exceptions, the process have been
# killed successfully, otherwise is still alive.
for counter in range(retries):
try:
_logger.debug("Checking with SIG_DFL if PID %s is still alive",
process_info.pid)
os.kill(process_info.pid, signal.SIG_DFL)
_logger.debug("os.kill call succeeded")
except OSError as e:
_logger.debug("os.kill call failed: %s", e)
# If the process doesn't exists, we are done.
if e.errno == errno.ESRCH:
return True
# Something unexpected has happened
output.error("%s", e)
return False
time.sleep(1)
_logger.debug("The PID %s has not been terminated after %s retries",
process_info.pid, retries)
return False
barman-2.1/barman/postgres.py 0000664 0001750 0001750 00000132574 13026732452 016712 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module represents the interface towards a PostgreSQL server.
"""
import atexit
import logging
from abc import ABCMeta
import psycopg2
from psycopg2.errorcodes import (DUPLICATE_OBJECT, OBJECT_IN_USE,
UNDEFINED_OBJECT)
from psycopg2.extensions import STATUS_IN_TRANSACTION
from psycopg2.extras import DictCursor, NamedTupleCursor
from barman.exceptions import (ConninfoException, PostgresAppNameError,
PostgresConnectionError,
PostgresDuplicateReplicationSlot,
PostgresException,
PostgresInvalidReplicationSlot,
PostgresIsInRecovery,
PostgresReplicationSlotInUse,
PostgresReplicationSlotsFull,
PostgresSuperuserRequired,
PostgresUnsupportedFeature)
from barman.infofile import Tablespace
from barman.remote_status import RemoteStatusMixin
from barman.utils import simplify_version, with_metaclass
# This is necessary because the CONFIGURATION_LIMIT_EXCEEDED constant
# has been added in psycopg2 2.5, but Barman supports version 2.4.2+ so
# in case of import error we declare a constant providing the correct value.
try:
from psycopg2.errorcodes import CONFIGURATION_LIMIT_EXCEEDED
except ImportError:
CONFIGURATION_LIMIT_EXCEEDED = '53400'
_logger = logging.getLogger(__name__)
_live_connections = []
"""
List of connections to be closed at the interpreter shutdown
"""
@atexit.register
def _atexit():
"""
Ensure that all the connections are correctly closed
at interpreter shutdown
"""
# Take a copy of the list because the conn.close() method modify it
for conn in list(_live_connections):
_logger.warn(
"Forcing %s cleanup during process shut down.",
conn.__class__.__name__)
conn.close()
class PostgreSQL(with_metaclass(ABCMeta, RemoteStatusMixin)):
"""
This abstract class represents a generic interface to a PostgreSQL server.
"""
def __init__(self, config, conninfo):
"""
Abstract base class constructor for PostgreSQL interface.
:param barman.config.ServerConfig config: the server configuration
:param str conninfo: Connection information (aka DSN)
"""
super(PostgreSQL, self).__init__()
assert conninfo
self.config = config
self.conninfo = conninfo
self._conn = None
# Build a dictionary with connection info parameters
# This is mainly used to speed up search in conninfo
try:
self.conn_parameters = self.parse_dsn(conninfo)
except (ValueError, TypeError) as e:
_logger.debug(e)
raise ConninfoException('Cannot connect to postgres: "%s" '
'is not a valid connection string' %
conninfo)
@staticmethod
def parse_dsn(dsn):
"""
Parse connection parameters from 'conninfo'
:param str dsn: Connection information (aka DSN)
:rtype: dict[str,str]
"""
# TODO: this might be made more robust in the future
return dict(x.split('=', 1) for x in dsn.split())
@staticmethod
def encode_dsn(parameters):
"""
Build a connection string from a dictionary of connection
parameters
:param dict[str,str] parameters: Connection parameters
:rtype: str
"""
# TODO: this might be made more robust in the future
return ' '.join(
["%s=%s" % (k, v) for k, v in sorted(parameters.items())])
def get_connection_string(self, application_name=None):
"""
Return the connection string, adding the application_name parameter
if requested, unless already defined by user in the connection string
:param str application_name: the application_name to add
:return str: the connection string
"""
conn_string = self.conninfo
# check if the application name is already defined by user
if application_name and 'application_name' not in self.conn_parameters:
# Then add the it to the connection string
conn_string += ' application_name=%s' % application_name
return conn_string
def connect(self):
"""
Generic function for Postgres connection (using psycopg2)
"""
if not self._conn:
try:
self._conn = psycopg2.connect(self.conninfo)
# If psycopg2 fails to connect to the host,
# raise the appropriate exception
except psycopg2.DatabaseError as e:
raise PostgresConnectionError(str(e).strip())
# Register the connection to the live connections list
_live_connections.append(self)
return self._conn
def close(self):
"""
Close the connection to PostgreSQL
"""
if self._conn:
if self._conn.status == STATUS_IN_TRANSACTION:
self._conn.rollback()
self._conn.close()
self._conn = None
# Remove the connection from the live connections list
_live_connections.remove(self)
def _cursor(self, *args, **kwargs):
"""
Return a cursor
"""
conn = self.connect()
return conn.cursor(*args, **kwargs)
@property
def server_version(self):
"""
Version of PostgreSQL (returned by psycopg2)
"""
conn = self.connect()
return conn.server_version
@property
def server_txt_version(self):
"""
Human readable version of PostgreSQL (calculated from server_version)
:rtype: str|None
"""
try:
conn = self.connect()
major = int(conn.server_version / 10000)
minor = int(conn.server_version / 100 % 100)
patch = int(conn.server_version % 100)
if major < 10:
return "%d.%d.%d" % (major, minor, patch)
if minor != 0:
_logger.warning(
"Unexpected non zero minor version %s in %s",
minor, conn.server_version)
return "%d.%d" % (major, patch)
except PostgresConnectionError as e:
_logger.debug("Error retrieving PostgreSQL version: %s",
str(e).strip())
return None
@property
def server_major_version(self):
"""
PostgreSQL major version (calculated from server_txt_version)
:rtype: str|None
"""
result = self.server_txt_version
if result is not None:
return simplify_version(result)
return None
class StreamingConnection(PostgreSQL):
"""
This class represents a streaming connection to a PostgreSQL server.
"""
def __init__(self, config):
"""
Streaming connection constructor
:param barman.config.ServerConfig config: the server configuration
"""
if config.streaming_conninfo is None:
raise ConninfoException(
"Missing 'streaming_conninfo' parameter for server '%s'"
% config.name)
super(StreamingConnection, self).__init__(config,
config.streaming_conninfo)
# Make sure we connect using the 'replication' option which
# triggers streaming replication protocol communication
self.conn_parameters['replication'] = 'true'
# Override 'dbname' parameter. This operation is required to mimic
# the behaviour of pg_receivexlog and pg_basebackup
self.conn_parameters['dbname'] = 'replication'
# Rebuild the conninfo string from the modified parameter lists
self.conninfo = self.encode_dsn(self.conn_parameters)
def connect(self):
"""
Connect to the PostgreSQL server. It reuses an existing connection.
:returns: the connection to the server
"""
if not self._conn:
# Build a connection and set autocommit
self._conn = super(StreamingConnection, self).connect()
self._conn.autocommit = True
return self._conn
def fetch_remote_status(self):
"""
Returns the status of the connection to the PostgreSQL server.
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
result = dict.fromkeys(
('connection_error', 'streaming_supported',
'streaming', 'systemid',
'timeline', 'xlogpos'),
None)
try:
# If the server is too old to support `pg_receivexlog`,
# exit immediately.
# This needs to be protected by the try/except because
# `self.server_version` can raise a PostgresConnectionError
if self.server_version < 90200:
result["streaming_supported"] = False
return result
result["streaming_supported"] = True
# Execute a IDENTIFY_SYSYEM to check the connection
cursor = self._cursor()
cursor.execute("IDENTIFY_SYSTEM")
row = cursor.fetchone()
# If something has been returned, barman is connected
# to a replication backend
if row:
result['streaming'] = True
# IDENTIFY_SYSTEM always return at least two values
result['systemid'] = row[0]
result['timeline'] = row[1]
# PostgreSQL 9.1+ returns also the current xlog flush location
if len(row) > 2:
result['xlogpos'] = row[2]
except psycopg2.ProgrammingError:
# This is not a streaming connection
result['streaming'] = False
except PostgresConnectionError as e:
result['connection_error'] = str(e).strip()
_logger.warn("Error retrieving PostgreSQL status: %s",
str(e).strip())
return result
def create_physical_repslot(self, slot_name):
"""
Create a physical replication slot using the streaming connection
:param str slot_name: Replication slot name
"""
cursor = self._cursor()
try:
# In the following query, the slot name is directly passed
# to the CREATE_REPLICATION_SLOT command, without any
# quoting. This is a characteristic of the streaming
# connection, otherwise if will fail with a generic
# "syntax error"
cursor.execute('CREATE_REPLICATION_SLOT %s PHYSICAL' % slot_name)
except psycopg2.DatabaseError as exc:
if exc.pgcode == DUPLICATE_OBJECT:
# A replication slot with the same name exists
raise PostgresDuplicateReplicationSlot()
elif exc.pgcode == CONFIGURATION_LIMIT_EXCEEDED:
# Unable to create a new physical replication slot.
# All slots are full.
raise PostgresReplicationSlotsFull()
else:
raise PostgresException(str(exc).strip())
def drop_repslot(self, slot_name):
"""
Drop a physical replication slot using the streaming connection
:param str slot_name: Replication slot name
"""
cursor = self._cursor()
try:
# In the following query, the slot name is directly passed
# to the DROP_REPLICATION_SLOT command, without any
# quoting. This is a characteristic of the streaming
# connection, otherwise if will fail with a generic
# "syntax error"
cursor.execute('DROP_REPLICATION_SLOT %s' % slot_name)
except psycopg2.DatabaseError as exc:
if exc.pgcode == UNDEFINED_OBJECT:
# A replication slot with the that name does not exist
raise PostgresInvalidReplicationSlot()
if exc.pgcode == OBJECT_IN_USE:
# The replication slot is still in use
raise PostgresReplicationSlotInUse()
else:
raise PostgresException(str(exc).strip())
class PostgreSQLConnection(PostgreSQL):
"""
This class represents a standard client connection to a PostgreSQL server.
"""
# Streaming replication client types
STANDBY = 1
WALSTREAMER = 2
ANY_STREAMING_CLIENT = (STANDBY, WALSTREAMER)
def __init__(self, config):
"""
PostgreSQL connection constructor.
:param barman.config.ServerConfig config: the server configuration
"""
# Check that 'conninfo' option is properly set
if config.conninfo is None:
raise ConninfoException(
"Missing 'conninfo' parameter for server '%s'" % config.name)
super(PostgreSQLConnection, self).__init__(config, config.conninfo)
self.configuration_files = None
def connect(self):
"""
Connect to the PostgreSQL server. It reuses an existing connection.
"""
if self._conn:
return self._conn
self._conn = super(PostgreSQLConnection, self).connect()
if (self._conn.server_version >= 90000 and
'application_name' not in self.conn_parameters):
try:
cur = self._conn.cursor()
cur.execute('SET application_name TO barman')
cur.close()
# If psycopg2 fails to set the application name,
# raise the appropriate exception
except psycopg2.ProgrammingError as e:
raise PostgresAppNameError(str(e).strip())
return self._conn
@property
def server_txt_version(self):
"""
Human readable version of PostgreSQL (returned by the server)
"""
try:
cur = self._cursor()
cur.execute("SELECT version()")
return cur.fetchone()[0].split()[1]
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving PostgreSQL version: %s",
str(e).strip())
return None
@property
def has_pgespresso(self):
"""
Returns true if the `pgespresso` extension is available
"""
try:
# pg_extension is only available from Postgres 9.1+
if self.server_version < 90100:
return False
cur = self._cursor()
cur.execute("SELECT count(*) FROM pg_extension "
"WHERE extname = 'pgespresso'")
q_result = cur.fetchone()[0]
return q_result > 0
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving pgespresso information: %s",
str(e).strip())
return None
@property
def is_in_recovery(self):
"""
Returns true if PostgreSQL server is in recovery mode (hot standby)
"""
try:
# pg_is_in_recovery is only available from Postgres 9.0+
if self.server_version < 90000:
return False
cur = self._cursor()
cur.execute("SELECT pg_is_in_recovery()")
return cur.fetchone()[0]
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error calling pg_is_in_recovery() function: %s",
str(e).strip())
return None
@property
def is_superuser(self):
"""
Returns true if current user has superuser privileges
"""
try:
cur = self._cursor()
cur.execute('SELECT usesuper FROM pg_user '
'WHERE usename = CURRENT_USER')
return cur.fetchone()[0]
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error calling is_superuser() function: %s",
str(e).strip())
return None
@property
def current_xlog_info(self):
"""
Get detailed information about the current WAL position in PostgreSQL.
This method returns a dictionary containing the following data:
* location
* file_name
* file_offset
* timestamp
When executed on a standby server file_name and file_offset are always
None
:rtype: psycopg2.extras.DictRow
"""
try:
cur = self._cursor(cursor_factory=DictCursor)
if not self.is_in_recovery:
cur.execute(
"SELECT location, "
"(pg_xlogfile_name_offset(location)).*, "
"CURRENT_TIMESTAMP AS timestamp "
"FROM pg_current_xlog_location() AS location")
return cur.fetchone()
else:
cur.execute(
"SELECT location, "
"NULL AS file_name, "
"NULL AS file_offset, "
"CURRENT_TIMESTAMP AS timestamp "
"FROM pg_last_xlog_replay_location() AS location")
return cur.fetchone()
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving current xlog "
"detailed information: %s",
str(e).strip())
return None
@property
def current_xlog_file_name(self):
"""
Get current WAL file from PostgreSQL
:return str: current WAL file in PostgreSQL
"""
current_xlog_info = self.current_xlog_info
if current_xlog_info is not None:
return current_xlog_info['file_name']
return None
@property
def current_xlog_location(self):
"""
Get current WAL location from PostgreSQL
:return str: current WAL location in PostgreSQL
"""
current_xlog_info = self.current_xlog_info
if current_xlog_info is not None:
return current_xlog_info['location']
return None
@property
def current_size(self):
"""
Returns the total size of the PostgreSQL server (requires superuser)
"""
if not self.is_superuser:
return None
try:
cur = self._cursor()
cur.execute(
"SELECT sum(pg_tablespace_size(oid)) "
"FROM pg_tablespace")
return cur.fetchone()[0]
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving PostgreSQL total size: %s",
str(e).strip())
return None
def get_archiver_stats(self):
"""
This method gathers statistics from pg_stat_archiver.
Only for Postgres 9.4+ or greater. If not available, returns None.
:return dict|None: a dictionary containing Postgres statistics from
pg_stat_archiver or None
"""
try:
# pg_stat_archiver is only available from Postgres 9.4+
if self.server_version < 90400:
return None
cur = self._cursor(cursor_factory=DictCursor)
# Select from pg_stat_archiver statistics view,
# retrieving statistics about WAL archiver process activity,
# also evaluating if the server is archiving without issues
# and the archived WALs per second rate.
#
# We are using current_settings to check for archive_mode=always.
# current_setting does normalise its output so we can just
# check for 'always' settings using a direct string
# comparison
cur.execute(
"SELECT *, "
"current_setting('archive_mode') IN ('on', 'always') "
"AND (last_failed_wal IS NULL "
"OR last_failed_wal LIKE '%.history' "
"AND substring(last_failed_wal from 1 for 8) "
"<= substring(last_archived_wal from 1 for 8) "
"OR last_failed_time <= last_archived_time) "
"AS is_archiving, "
"CAST (archived_count AS NUMERIC) "
"/ EXTRACT (EPOCH FROM age(now(), stats_reset)) "
"AS current_archived_wals_per_second "
"FROM pg_stat_archiver")
return cur.fetchone()
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving pg_stat_archive data: %s",
str(e).strip())
return None
def fetch_remote_status(self):
"""
Get the status of the PostgreSQL server
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
# PostgreSQL settings to get from the server (requiring superuser)
pg_superuser_settings = [
'data_directory']
# PostgreSQL settings to get from the server
pg_settings = []
pg_query_keys = [
'server_txt_version',
'is_superuser',
'current_xlog',
'pgespresso_installed',
'replication_slot_support',
'replication_slot',
'synchronous_standby_names',
]
# Initialise the result dictionary setting all the values to None
result = dict.fromkeys(pg_superuser_settings +
pg_settings +
pg_query_keys,
None)
try:
# check for wal_level only if the version is >= 9.0
if self.server_version >= 90000:
pg_settings.append('wal_level')
# retrieves superuser settings
if self.is_superuser:
for name in pg_superuser_settings:
result[name] = self.get_setting(name)
# retrieves standard settings
for name in pg_settings:
result[name] = self.get_setting(name)
result['is_superuser'] = self.is_superuser
result['server_txt_version'] = self.server_txt_version
result['pgespresso_installed'] = self.has_pgespresso
result['current_xlog'] = self.current_xlog_file_name
result['current_size'] = self.current_size
result.update(self.get_configuration_files())
# Retrieve the replication_slot status
result["replication_slot_support"] = False
if self.server_version >= 90400:
result["replication_slot_support"] = True
if self.config.slot_name is not None:
result["replication_slot"] = (
self.get_replication_slot(self.config.slot_name))
# Retrieve the list of synchronous standby names
result["synchronous_standby_names"] = []
if self.server_version >= 90100:
result["synchronous_standby_names"] = (
self.get_synchronous_standby_names())
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.warn("Error retrieving PostgreSQL status: %s",
str(e).strip())
return result
def get_setting(self, name):
"""
Get a Postgres setting with a given name
:param name: a parameter name
"""
try:
cur = self._cursor()
cur.execute('SHOW "%s"' % name.replace('"', '""'))
return cur.fetchone()[0]
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving PostgreSQL setting '%s': %s",
name.replace('"', '""'), str(e).strip())
return None
def get_tablespaces(self):
"""
Returns a list of tablespaces or None if not present
"""
try:
cur = self._cursor()
if self.server_version >= 90200:
cur.execute(
"SELECT spcname, oid, "
"pg_tablespace_location(oid) AS spclocation "
"FROM pg_tablespace "
"WHERE pg_tablespace_location(oid) != ''")
else:
cur.execute(
"SELECT spcname, oid, spclocation "
"FROM pg_tablespace WHERE spclocation != ''")
# Generate a list of tablespace objects
return [Tablespace._make(item) for item in cur.fetchall()]
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving PostgreSQL tablespaces: %s",
str(e).strip())
return None
def get_configuration_files(self):
"""
Get postgres configuration files or an empty dictionary
in case of error
:rtype: dict
"""
if self.configuration_files:
return self.configuration_files
try:
self.configuration_files = {}
cur = self._cursor()
cur.execute(
"SELECT name, setting FROM pg_settings "
"WHERE name IN ('config_file', 'hba_file', 'ident_file')")
for cname, cpath in cur.fetchall():
self.configuration_files[cname] = cpath
# Retrieve additional configuration files
# If PostgresSQL is older than 8.4 disable this check
if self.server_version >= 80400:
cur.execute(
"SELECT DISTINCT sourcefile AS included_file "
"FROM pg_settings "
"WHERE sourcefile IS NOT NULL "
"AND sourcefile NOT IN "
"(SELECT setting FROM pg_settings "
"WHERE name = 'config_file') "
"ORDER BY 1")
# Extract the values from the containing single element tuples
included_files = [included_file
for included_file, in cur.fetchall()]
if len(included_files) > 0:
self.configuration_files['included_files'] = included_files
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving PostgreSQL configuration files "
"location: %s", str(e).strip())
self.configuration_files = {}
return self.configuration_files
def create_restore_point(self, target_name):
"""
Create a restore point with the given target name
The method executes the pg_create_restore_point() function through
a PostgreSQL connection. Only for Postgres versions >= 9.1 when not
in replication.
If requirements are not met, the operation is skipped.
:param str target_name: name of the restore point
:returns: the restore point LSN
:rtype: str|None
"""
if self.server_version < 90100:
return None
# Not possible if on a standby
# Called inside the pg_connect context to reuse the connection
if self.is_in_recovery:
return None
try:
cur = self._cursor()
cur.execute(
"SELECT pg_create_restore_point(%s)", [target_name])
return cur.fetchone()[0]
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug('Error issuing pg_create_restore_point()'
'command: %s', str(e).strip())
return None
def start_exclusive_backup(self, label):
"""
Calls pg_start_backup() on the PostgreSQL server
This method returns a dictionary containing the following data:
* location
* file_name
* file_offset
* timestamp
:param str label: descriptive string to identify the backup
:rtype: psycopg2.extras.DictRow
"""
try:
conn = self.connect()
# Rollback to release the transaction, as the pg_start_backup
# invocation can last up to PostgreSQL's checkpoint_timeout
conn.rollback()
# Start an exclusive backup
cur = conn.cursor(cursor_factory=DictCursor)
if self.server_version < 80400:
cur.execute(
"SELECT location, "
"(pg_xlogfile_name_offset(location)).*, "
"now() AS timestamp "
"FROM pg_start_backup(%s) AS location",
(label,))
else:
cur.execute(
"SELECT location, "
"(pg_xlogfile_name_offset(location)).*, "
"now() AS timestamp "
"FROM pg_start_backup(%s,%s) AS location",
(label, self.config.immediate_checkpoint))
start_row = cur.fetchone()
# Rollback to release the transaction, as the connection
# is to be retained until the end of backup
conn.rollback()
return start_row
except (PostgresConnectionError, psycopg2.Error) as e:
msg = "pg_start_backup(): %s" % str(e).strip()
_logger.debug(msg)
raise PostgresException(msg)
def start_concurrent_backup(self, label):
"""
Calls pg_start_backup on the PostgreSQL server using the
API introduced with version 9.6
This method returns a dictionary containing the following data:
* location
* timeline
* timestamp
:param str label: descriptive string to identify the backup
:rtype: psycopg2.extras.DictRow
"""
try:
conn = self.connect()
# Rollback to release the transaction, as the pg_start_backup
# invocation can last up to PostgreSQL's checkpoint_timeout
conn.rollback()
# Start the backup using the api introduced in postgres 9.6
cur = conn.cursor(cursor_factory=DictCursor)
cur.execute(
"SELECT location, "
"(SELECT timeline_id "
"FROM pg_control_checkpoint()) AS timeline, "
"now() AS timestamp "
"FROM pg_start_backup(%s, %s, FALSE) AS location",
(label, self.config.immediate_checkpoint))
start_row = cur.fetchone()
# Rollback to release the transaction, as the connection
# is to be retained until the end of backup
conn.rollback()
return start_row
except (PostgresConnectionError, psycopg2.Error) as e:
msg = "pg_start_backup command: %s" % (str(e).strip(),)
_logger.debug(msg)
raise PostgresException(msg)
def stop_exclusive_backup(self):
"""
Calls pg_stop_backup() on the PostgreSQL server
This method returns a dictionary containing the following data:
* location
* file_name
* file_offset
* timestamp
:rtype: psycopg2.extras.DictRow
"""
try:
conn = self.connect()
# Rollback to release the transaction, as the pg_stop_backup
# invocation could will wait until the current WAL file is shipped
conn.rollback()
# Stop the backup
cur = conn.cursor(cursor_factory=DictCursor)
cur.execute(
"SELECT location, "
"(pg_xlogfile_name_offset(location)).*, "
"now() AS timestamp "
"FROM pg_stop_backup() AS location")
return cur.fetchone()
except (PostgresConnectionError, psycopg2.Error) as e:
msg = "Error issuing pg_stop_backup command: %s" % str(e).strip()
_logger.debug(msg)
raise PostgresException(
'Cannot terminate exclusive backup. '
'You might have to manually execute pg_stop_backup '
'on your PostgreSQL server')
def stop_concurrent_backup(self):
"""
Calls pg_stop_backup on the PostgreSQL server using the
API introduced with version 9.6
This method returns a dictionary containing the following data:
* location
* timeline
* backup_label
* timestamp
:rtype: psycopg2.extras.DictRow
"""
try:
conn = self.connect()
# Rollback to release the transaction, as the pg_stop_backup
# invocation could will wait until the current WAL file is shipped
conn.rollback()
# Stop the backup using the api introduced with version 9.6
cur = conn.cursor(cursor_factory=DictCursor)
cur.execute(
'SELECT end_row.lsn AS location, '
'(SELECT CASE WHEN pg_is_in_recovery() '
'THEN min_recovery_end_timeline ELSE timeline_id END '
'FROM pg_control_checkpoint(), pg_control_recovery()'
') AS timeline, '
'end_row.labelfile AS backup_label, '
'now() AS timestamp FROM pg_stop_backup(FALSE) AS end_row')
return cur.fetchone()
except (PostgresConnectionError, psycopg2.Error) as e:
msg = "Error issuing pg_stop_backup command: %s" % str(e).strip()
_logger.debug(msg)
raise PostgresException(msg)
def pgespresso_start_backup(self, label):
"""
Execute a pgespresso_start_backup
This method returns a dictionary containing the following data:
* backup_label
* timestamp
:param str label: descriptive string to identify the backup
:rtype: psycopg2.extras.DictRow
"""
try:
conn = self.connect()
# Rollback to release the transaction,
# as the pgespresso_start_backup invocation can last
# up to PostgreSQL's checkpoint_timeout
conn.rollback()
# Start the concurrent backup using pgespresso
cur = conn.cursor(cursor_factory=DictCursor)
cur.execute(
'SELECT pgespresso_start_backup(%s,%s) AS backup_label, '
'now() AS timestamp',
(label, self.config.immediate_checkpoint))
start_row = cur.fetchone()
# Rollback to release the transaction, as the connection
# is to be retained until the end of backup
conn.rollback()
return start_row
except (PostgresConnectionError, psycopg2.Error) as e:
msg = "pgespresso_start_backup(): %s" % str(e).strip()
_logger.debug(msg)
raise PostgresException(msg)
def pgespresso_stop_backup(self, backup_label):
"""
Execute a pgespresso_stop_backup
This method returns a dictionary containing the following data:
* end_wal
* timestamp
:param str backup_label: backup label as returned
by pgespress_start_backup
:rtype: psycopg2.extras.DictRow
"""
try:
conn = self.connect()
# Issue a rollback to release any unneeded lock
conn.rollback()
cur = conn.cursor(cursor_factory=DictCursor)
cur.execute("SELECT pgespresso_stop_backup(%s) AS end_wal, "
"now() AS timestamp",
(backup_label,))
return cur.fetchone()
except (PostgresConnectionError, psycopg2.Error) as e:
msg = "Error issuing pgespresso_stop_backup() command: %s" % (
str(e).strip())
_logger.debug(msg)
raise PostgresException(
'%s\n'
'HINT: You might have to manually execute '
'pgespresso_abort_backup() on your PostgreSQL '
'server' % msg)
def switch_xlog(self):
"""
Execute a pg_switch_xlog()
To be SURE of the switch of a xlog, we collect the xlogfile name
before and after the switch.
The method returns the just closed xlog file name if the current xlog
file has changed, it returns an empty string otherwise.
The method returns None if something went wrong during the execution
of the pg_switch_xlog command.
:rtype: str|None
"""
try:
conn = self.connect()
# Requires superuser privilege
if not self.is_superuser:
raise PostgresSuperuserRequired()
# If this server is in recovery there is nothing to do
if self.is_in_recovery:
raise PostgresIsInRecovery()
cur = conn.cursor()
# Collect the xlog file name before the switch
cur.execute('SELECT pg_xlogfile_name('
'pg_current_xlog_insert_location())')
pre_switch = cur.fetchone()[0]
# Switch
cur.execute('SELECT pg_xlogfile_name(pg_switch_xlog())')
# Collect the xlog file name after the switch
cur.execute('SELECT pg_xlogfile_name('
'pg_current_xlog_insert_location())')
post_switch = cur.fetchone()[0]
if pre_switch < post_switch:
return pre_switch
else:
return ''
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug(
"Error issuing pg_switch_xlog() command: %s",
str(e).strip())
return None
def checkpoint(self):
"""
Execute a checkpoint
"""
try:
conn = self.connect()
# Requires superuser privilege
if not self.is_superuser:
raise PostgresSuperuserRequired()
cur = conn.cursor()
cur.execute("CHECKPOINT")
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug(
"Error issuing CHECKPOINT: %s",
str(e).strip())
def get_replication_stats(self, client_type=STANDBY):
"""
Returns streaming replication information
"""
try:
cur = self._cursor(cursor_factory=NamedTupleCursor)
# Without superuser rights, this function is useless
# TODO: provide a simplified version for non-superusers
if not self.is_superuser:
raise PostgresSuperuserRequired()
# pg_stat_replication is a system view that contains one
# row per WAL sender process with information about the
# replication status of a standby server. It has been
# introduced in PostgreSQL 9.1. Current fields are:
#
# - pid (procpid in 9.1)
# - usesysid
# - usename
# - application_name
# - client_addr
# - client_hostname
# - client_port
# - backend_start
# - backend_xmin (9.4+)
# - state
# - sent_location
# - write_location
# - flush_location
# - replay_location
# - sync_priority
# - sync_state
#
if self.server_version < 90100:
raise PostgresUnsupportedFeature('9.1')
from_repslot = ""
if self.server_version >= 90500:
# Current implementation (9.5+)
what = "r.*, rs.slot_name"
# Look for replication slot name
from_repslot = "LEFT JOIN pg_replication_slots rs " \
"ON (r.pid = rs.active_pid) "
elif self.server_version >= 90400:
# PostgreSQL 9.4
what = "*"
elif self.server_version >= 90200:
# PostgreSQL 9.2/9.3
what = "pid," \
"usesysid," \
"usename," \
"application_name," \
"client_addr," \
"client_hostname," \
"client_port," \
"backend_start," \
"CAST (NULL AS xid) AS backend_xmin," \
"state," \
"sent_location," \
"write_location," \
"flush_location," \
"replay_location," \
"sync_priority," \
"sync_state "
else:
# PostgreSQL 9.1
what = "procpid AS pid," \
"usesysid," \
"usename," \
"application_name," \
"client_addr," \
"client_hostname," \
"client_port," \
"backend_start," \
"CAST (NULL AS xid) AS backend_xmin," \
"state," \
"sent_location," \
"write_location," \
"flush_location," \
"replay_location," \
"sync_priority," \
"sync_state "
# Streaming client
if client_type == self.STANDBY:
# Standby server
where = 'WHERE replay_location IS NOT NULL '
elif client_type == self.WALSTREAMER:
# WAL streamer
where = 'WHERE replay_location IS NULL '
else:
where = ''
# Execute the query
cur.execute(
"SELECT %s, "
"pg_is_in_recovery() AS is_in_recovery,"
"CASE WHEN pg_is_in_recovery() "
" THEN NULL "
" ELSE pg_current_xlog_location() "
"END AS current_location "
"FROM pg_stat_replication r "
"%s"
"%s"
"ORDER BY sync_state DESC, sync_priority" % (what,
from_repslot,
where))
# Generate a list of standby objects
return cur.fetchall()
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving status of standby servers: %s",
str(e).strip())
return None
def get_replication_slot(self, slot_name):
"""
Retrieve from the PostgreSQL server a physical replication slot
with a specific slot_name.
This method returns a dictionary containing the following data:
* slot_name
* active
* restart_lsn
:param str slot_name: the replication slot name
:rtype: psycopg2.extras.DictRow
"""
if self.server_version < 90400:
# Raise exception if replication slot are not supported
# by PostgreSQL version
raise PostgresUnsupportedFeature('9.4')
else:
cur = self._cursor(cursor_factory=NamedTupleCursor)
try:
cur.execute("SELECT slot_name, "
"active, "
"restart_lsn "
"FROM pg_replication_slots "
"WHERE slot_type = 'physical' "
"AND slot_name = '%s'" % slot_name)
# Retrieve the replication slot information
return cur.fetchone()
except (PostgresConnectionError, psycopg2.Error) as e:
_logger.debug("Error retrieving replication_slots: %s",
str(e).strip())
raise
def get_synchronous_standby_names(self):
"""
Retrieve the list of named synchronous standby servers from PostgreSQL
This method returns a list of names
:return list: synchronous standby names
"""
if self.server_version < 90100:
# Raise exception if synchronous replication is not supported
raise PostgresUnsupportedFeature('9.1')
else:
synchronous_standby_names = (
self.get_setting('synchronous_standby_names'))
# Normalise the list of sync standby names
# On PostgreSQL 9.6 it is possible to specify the number of
# required synchronous standby using this format:
# n (name1, name2, ... nameN).
# We only need the name list, so we discard everything else.
# The name list starts after the first parenthesis or at pos 0
names_start = synchronous_standby_names.find('(') + 1
names_end = synchronous_standby_names.rfind(')')
if names_end < 0:
names_end = len(synchronous_standby_names)
names_list = synchronous_standby_names[names_start:names_end]
return [x.strip() for x in names_list.split(',')]
barman-2.1/barman/backup.py 0000664 0001750 0001750 00000111222 13033425127 016270 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module represents a backup.
"""
import datetime
import logging
import os
import shutil
from glob import glob
import dateutil.parser
import dateutil.tz
from barman import output, xlog
from barman.backup_executor import PostgresBackupExecutor, RsyncBackupExecutor
from barman.compression import CompressionManager
from barman.config import BackupOptions
from barman.exceptions import (AbortedRetryHookScript,
CompressionIncompatibility, SshCommandException,
UnknownBackupIdException)
from barman.hooks import HookScriptRunner, RetryHookScriptRunner
from barman.infofile import BackupInfo, WalFileInfo
from barman.recovery_executor import RecoveryExecutor
from barman.remote_status import RemoteStatusMixin
from barman.utils import fsync_dir, human_readable_timedelta, pretty_size
_logger = logging.getLogger(__name__)
class BackupManager(RemoteStatusMixin):
"""Manager of the backup archive for a server"""
DEFAULT_STATUS_FILTER = (BackupInfo.DONE,)
def __init__(self, server):
"""
Constructor
"""
super(BackupManager, self).__init__()
self.server = server
self.config = server.config
self._backup_cache = None
self.compression_manager = CompressionManager(self.config, server.path)
self.executor = None
try:
if self.config.backup_method == "postgres":
self.executor = PostgresBackupExecutor(self)
else:
self.executor = RsyncBackupExecutor(self)
except SshCommandException as e:
self.config.disabled = True
self.config.msg_list.append(str(e).strip())
@property
def mode(self):
"""
Property defining the BackupInfo mode content
"""
if self.executor:
return self.executor.mode
return None
def get_available_backups(self, status_filter=DEFAULT_STATUS_FILTER):
"""
Get a list of available backups
:param status_filter: default DEFAULT_STATUS_FILTER. The status of
the backup list returned
"""
# If the filter is not a tuple, create a tuple using the filter
if not isinstance(status_filter, tuple):
status_filter = tuple(status_filter,)
# Load the cache if necessary
if self._backup_cache is None:
self._load_backup_cache()
# Filter the cache using the status filter tuple
backups = {}
for key, value in self._backup_cache.items():
if value.status in status_filter:
backups[key] = value
return backups
def _load_backup_cache(self):
"""
Populate the cache of the available backups, reading information
from disk.
"""
self._backup_cache = {}
# Load all the backups from disk reading the backup.info files
for filename in glob("%s/*/backup.info" %
self.config.basebackups_directory):
backup = BackupInfo(self.server, filename)
self._backup_cache[backup.backup_id] = backup
def backup_cache_add(self, backup_info):
"""
Register a BackupInfo object to the backup cache.
NOTE: Initialise the cache - in case it has not been done yet
:param barman.infofile.BackupInfo backup_info: the object we want to
register in the cache
"""
# Load the cache if needed
if self._backup_cache is None:
self._load_backup_cache()
# Insert the BackupInfo object into the cache
self._backup_cache[backup_info.backup_id] = backup_info
def backup_cache_remove(self, backup_info):
"""
Remove a BackupInfo object from the backup cache
This method _must_ be called after removing the object from disk.
:param barman.infofile.BackupInfo backup_info: the object we want to
remove from the cache
"""
# Nothing to do if the cache is not loaded
if self._backup_cache is None:
return
# Remove the BackupInfo object from the backups cache
del self._backup_cache[backup_info.backup_id]
def get_backup(self, backup_id):
"""
Return the backup information for the given backup id.
If the backup_id is None or backup.info file doesn't exists,
it returns None.
:param str|None backup_id: the ID of the backup to return
:rtype: BackupInfo|None
"""
if backup_id is not None:
# Get all the available backups from the cache
available_backups = self.get_available_backups(
BackupInfo.STATUS_ALL)
# Return the BackupInfo if present, or None
return available_backups.get(backup_id)
return None
def get_previous_backup(self, backup_id,
status_filter=DEFAULT_STATUS_FILTER):
"""
Get the previous backup (if any) in the catalog
:param status_filter: default DEFAULT_STATUS_FILTER. The status of
the backup returned
"""
if not isinstance(status_filter, tuple):
status_filter = tuple(status_filter)
backup = BackupInfo(self.server, backup_id=backup_id)
available_backups = self.get_available_backups(status_filter +
(backup.status,))
ids = sorted(available_backups.keys())
try:
current = ids.index(backup_id)
while current > 0:
res = available_backups[ids[current - 1]]
if res.status in status_filter:
return res
current -= 1
return None
except ValueError:
raise UnknownBackupIdException('Could not find backup_id %s' %
backup_id)
def get_next_backup(self, backup_id, status_filter=DEFAULT_STATUS_FILTER):
"""
Get the next backup (if any) in the catalog
:param status_filter: default DEFAULT_STATUS_FILTER. The status of
the backup returned
"""
if not isinstance(status_filter, tuple):
status_filter = tuple(status_filter)
backup = BackupInfo(self.server, backup_id=backup_id)
available_backups = self.get_available_backups(status_filter +
(backup.status,))
ids = sorted(available_backups.keys())
try:
current = ids.index(backup_id)
while current < (len(ids) - 1):
res = available_backups[ids[current + 1]]
if res.status in status_filter:
return res
current += 1
return None
except ValueError:
raise UnknownBackupIdException('Could not find backup_id %s' %
backup_id)
def get_last_backup_id(self, status_filter=DEFAULT_STATUS_FILTER):
"""
Get the id of the latest/last backup in the catalog (if exists)
:param status_filter: The status of the backup to return,
default to DEFAULT_STATUS_FILTER.
:return string|None: ID of the backup
"""
available_backups = self.get_available_backups(status_filter)
if len(available_backups) == 0:
return None
ids = sorted(available_backups.keys())
return ids[-1]
def get_first_backup_id(self, status_filter=DEFAULT_STATUS_FILTER):
"""
Get the id of the oldest/first backup in the catalog (if exists)
:param status_filter: The status of the backup to return,
default to DEFAULT_STATUS_FILTER.
:return string|None: ID of the backup
"""
available_backups = self.get_available_backups(status_filter)
if len(available_backups) == 0:
return None
ids = sorted(available_backups.keys())
return ids[0]
def delete_backup(self, backup):
"""
Delete a backup
:param backup: the backup to delete
"""
available_backups = self.get_available_backups()
minimum_redundancy = self.server.config.minimum_redundancy
# Honour minimum required redundancy
if backup.status == BackupInfo.DONE and \
minimum_redundancy >= len(available_backups):
output.warning("Skipping delete of backup %s for server %s "
"due to minimum redundancy requirements "
"(minimum redundancy = %s, "
"current redundancy = %s)",
backup.backup_id,
self.config.name,
len(available_backups),
minimum_redundancy)
return
# Keep track of when the delete operation started.
delete_start_time = datetime.datetime.now()
output.info("Deleting backup %s for server %s",
backup.backup_id, self.config.name)
previous_backup = self.get_previous_backup(backup.backup_id)
next_backup = self.get_next_backup(backup.backup_id)
# Delete all the data contained in the backup
try:
self.delete_backup_data(backup)
except OSError as e:
output.error("Failure deleting backup %s for server %s.\n%s",
backup.backup_id, self.config.name, e)
return
# Check if we are deleting the first available backup
if not previous_backup:
# In the case of exclusive backup (default), removes any WAL
# files associated to the backup being deleted.
# In the case of concurrent backup, removes only WAL files
# prior to the start of the backup being deleted, as they
# might be useful to any concurrent backup started immediately
# after.
remove_until = None # means to remove all WAL files
if next_backup:
remove_until = next_backup
elif BackupOptions.CONCURRENT_BACKUP in self.config.backup_options:
remove_until = backup
timelines_to_protect = set()
# If remove_until is not set there are no backup left
if remove_until:
# Retrieve the list of extra timelines that contains at least
# a backup. On such timelines we don't want to delete any WAL
for value in self.get_available_backups(
BackupInfo.STATUS_ARCHIVING).values():
# Ignore the backup that is being deleted
if value == backup:
continue
timelines_to_protect.add(value.timeline)
# Remove the timeline of `remove_until` from the list.
# We have enough information to safely delete unused WAL files
# on it.
timelines_to_protect -= set([remove_until.timeline])
output.info("Delete associated WAL segments:")
for name in self.remove_wal_before_backup(remove_until,
timelines_to_protect):
output.info("\t%s", name)
# As last action, remove the backup directory,
# ending the delete operation
try:
self.delete_basebackup(backup)
except OSError as e:
output.error("Failure deleting backup %s for server %s.\n%s\n"
"Please manually remove the '%s' directory",
backup.backup_id, self.config.name, e,
backup.get_basebackup_directory())
return
self.backup_cache_remove(backup)
# Save the time of the complete removal of the backup
delete_end_time = datetime.datetime.now()
output.info("Deleted backup %s (start time: %s, elapsed time: %s)",
backup.backup_id,
delete_start_time.ctime(),
human_readable_timedelta(
delete_end_time - delete_start_time))
def backup(self):
"""
Performs a backup for the server
"""
_logger.debug("initialising backup information")
self.executor.init()
backup_info = None
try:
# Create the BackupInfo object representing the backup
backup_info = BackupInfo(
self.server,
backup_id=datetime.datetime.now().strftime('%Y%m%dT%H%M%S'))
backup_info.save()
self.backup_cache_add(backup_info)
output.info(
"Starting backup using %s method for server %s in %s",
self.mode,
self.config.name,
backup_info.get_basebackup_directory())
# Run the pre-backup-script if present.
script = HookScriptRunner(self, 'backup_script', 'pre')
script.env_from_backup_info(backup_info)
script.run()
# Run the pre-backup-retry-script if present.
retry_script = RetryHookScriptRunner(
self, 'backup_retry_script', 'pre')
retry_script.env_from_backup_info(backup_info)
retry_script.run()
# Do the backup using the BackupExecutor
self.executor.backup(backup_info)
# Compute backup size and fsync it on disk
self.backup_fsync_and_set_sizes(backup_info)
# Mark the backup as DONE
backup_info.set_attribute("status", "DONE")
# Use BaseException instead of Exception to catch events like
# KeyboardInterrupt (e.g.: CRTL-C)
except BaseException as e:
msg_lines = str(e).strip().splitlines()
if backup_info:
# Use only the first line of exception message
# in backup_info error field
backup_info.set_attribute("status", "FAILED")
# If the exception has no attached message use the raw
# type name
if len(msg_lines) == 0:
msg_lines = [type(e).__name__]
backup_info.set_attribute(
"error",
"failure %s (%s)" % (
self.executor.current_action, msg_lines[0]))
output.error("Backup failed %s.\nDETAILS: %s\n%s",
self.executor.current_action, msg_lines[0],
'\n'.join(msg_lines[1:]))
else:
output.info("Backup end at xlog location: %s (%s, %08X)",
backup_info.end_xlog,
backup_info.end_wal,
backup_info.end_offset)
output.info("Backup completed")
# Create a restore point after a backup
target_name = 'barman_%s' % backup_info.backup_id
self.server.postgres.create_restore_point(target_name)
finally:
if backup_info:
backup_info.save()
# Make sure we are not holding any PostgreSQL connection
# during the post-backup scripts
self.server.close()
# Run the post-backup-retry-script if present.
try:
retry_script = RetryHookScriptRunner(
self, 'backup_retry_script', 'post')
retry_script.env_from_backup_info(backup_info)
retry_script.run()
except AbortedRetryHookScript as e:
# Ignore the ABORT_STOP as it is a post-hook operation
_logger.warning("Ignoring stop request after receiving "
"abort (exit code %d) from post-backup "
"retry hook script: %s",
e.hook.exit_status, e.hook.script)
# Run the post-backup-script if present.
script = HookScriptRunner(self, 'backup_script', 'post')
script.env_from_backup_info(backup_info)
script.run()
output.result('backup', backup_info)
def recover(self, backup_info, dest, tablespaces=None, target_tli=None,
target_time=None, target_xid=None, target_name=None,
exclusive=False, remote_command=None):
"""
Performs a recovery of a backup
:param barman.infofile.BackupInfo backup_info: the backup to recover
:param str dest: the destination directory
:param dict[str,str]|None tablespaces: a tablespace name -> location
map (for relocation)
:param str|None target_tli: the target timeline
:param str|None target_time: the target time
:param str|None target_xid: the target xid
:param str|None target_name: the target name created previously with
pg_create_restore_point() function call
:param bool exclusive: whether the recovery is exclusive or not
:param str|None remote_command: default None. The remote command
to recover the base backup, in case of remote backup.
"""
# Archive every WAL files in the incoming directory of the server
self.server.archive_wal(verbose=False)
# Delegate the recovery operation to a RecoveryExecutor object
executor = RecoveryExecutor(self)
recovery_info = executor.recover(backup_info,
dest, tablespaces,
target_tli, target_time,
target_xid, target_name,
exclusive, remote_command)
# Output recovery results
output.result('recovery', recovery_info['results'])
def archive_wal(self, verbose=True):
"""
Executes WAL maintenance operations, such as archiving and compression
If verbose is set to False, outputs something only if there is
at least one file
:param bool verbose: report even if no actions
"""
with self.server.xlogdb('a') as fxlogdb:
for archiver in self.server.archivers:
archiver.archive(fxlogdb, verbose)
def cron_retention_policy(self):
"""
Retention policy management
"""
if (self.server.enforce_retention_policies and
self.config.retention_policy_mode == 'auto'):
available_backups = self.get_available_backups(
BackupInfo.STATUS_ALL)
retention_status = self.config.retention_policy.report()
for bid in sorted(retention_status.keys()):
if retention_status[bid] == BackupInfo.OBSOLETE:
output.info(
"Enforcing retention policy: removing backup %s for "
"server %s" % (bid, self.config.name))
self.delete_backup(available_backups[bid])
def delete_basebackup(self, backup):
"""
Delete the basebackup dir of a given backup.
:param barman.infofile.BackupInfo backup: the backup to delete
"""
backup_dir = backup.get_basebackup_directory()
_logger.debug("Deleting base backup directory: %s" % backup_dir)
shutil.rmtree(backup_dir)
def delete_backup_data(self, backup):
"""
Delete the data contained in a given backup.
:param barman.infofile.BackupInfo backup: the backup to delete
"""
if backup.tablespaces:
if backup.backup_version == 2:
tbs_dir = backup.get_basebackup_directory()
else:
tbs_dir = os.path.join(backup.get_data_directory(),
'pg_tblspc')
for tablespace in backup.tablespaces:
rm_dir = os.path.join(tbs_dir, str(tablespace.oid))
if os.path.exists(rm_dir):
_logger.debug("Deleting tablespace %s directory: %s" %
(tablespace.name, rm_dir))
shutil.rmtree(rm_dir)
pg_data = backup.get_data_directory()
if os.path.exists(pg_data):
_logger.debug("Deleting PGDATA directory: %s" % pg_data)
shutil.rmtree(pg_data)
def delete_wal(self, wal_info):
"""
Delete a WAL segment, with the given WalFileInfo
:param barman.infofile.WalFileInfo wal_info: the WAL to delete
"""
try:
os.unlink(wal_info.fullpath(self.server))
try:
os.removedirs(os.path.dirname(wal_info.fullpath(self.server)))
except OSError:
# This is not an error condition
# We always try to remove the the trailing directories,
# this means that hashdir is not empty.
pass
except OSError as e:
output.warning('Ignoring deletion of WAL file %s '
'for server %s: %s',
wal_info.name, self.config.name, e)
def check(self, check_strategy):
"""
This function does some checks on the server.
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
# Check compression_setting parameter
if self.config.compression and not self.compression_manager.check():
check_strategy.result(self.config.name,
'compression settings', False)
else:
status = True
try:
self.compression_manager.get_compressor()
except CompressionIncompatibility as field:
check_strategy.result(self.config.name,
'%s setting' % field, False)
status = False
check_strategy.result(self.config.name,
'compression settings', status)
# Failed backups check
failed_backups = self.get_available_backups((BackupInfo.FAILED,))
status = len(failed_backups) == 0
check_strategy.result(
self.config.name,
'failed backups',
status,
'there are %s failed backups' % (len(failed_backups,))
)
# Minimum redundancy checks
no_backups = len(self.get_available_backups())
# Check minimum_redundancy_requirements parameter
if no_backups < int(self.config.minimum_redundancy):
status = False
else:
status = True
check_strategy.result(
self.config.name,
'minimum redundancy requirements', status,
'have %s backups, expected at least %s' % (
no_backups, self.config.minimum_redundancy))
# TODO: Add a check for the existence of ssh and of rsync
# Execute additional checks defined by the BackupExecutor
if self.executor:
self.executor.check(check_strategy)
def status(self):
"""
This function show the server status
"""
# get number of backups
no_backups = len(self.get_available_backups())
output.result('status', self.config.name,
"backups_number",
"No. of available backups", no_backups)
output.result('status', self.config.name,
"first_backup",
"First available backup",
self.get_first_backup_id())
output.result('status', self.config.name,
"last_backup",
"Last available backup",
self.get_last_backup_id())
# Minimum redundancy check. if number of backups minor than minimum
# redundancy, fail.
if no_backups < self.config.minimum_redundancy:
output.result('status', self.config.name,
"minimum_redundancy",
"Minimum redundancy requirements",
"FAILED (%s/%s)" % (
no_backups,
self.config.minimum_redundancy))
else:
output.result('status', self.config.name,
"minimum_redundancy",
"Minimum redundancy requirements",
"satisfied (%s/%s)" % (
no_backups,
self.config.minimum_redundancy))
# Output additional status defined by the BackupExecutor
if self.executor:
self.executor.status()
def fetch_remote_status(self):
"""
Build additional remote status lines defined by the BackupManager.
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
if self.executor:
return self.executor.get_remote_status()
else:
return {}
def rebuild_xlogdb(self):
"""
Rebuild the whole xlog database guessing it from the archive content.
"""
from os.path import isdir, join
output.info("Rebuilding xlogdb for server %s", self.config.name)
root = self.config.wals_directory
default_compression = self.config.compression
wal_count = label_count = history_count = 0
# lock the xlogdb as we are about replacing it completely
with self.server.xlogdb('w') as fxlogdb:
xlogdb_new = fxlogdb.name + ".new"
with open(xlogdb_new, 'w') as fxlogdb_new:
for name in sorted(os.listdir(root)):
# ignore the xlogdb and its lockfile
if name.startswith(self.server.XLOG_DB):
continue
fullname = join(root, name)
if isdir(fullname):
# all relevant files are in subdirectories
hash_dir = fullname
for wal_name in sorted(os.listdir(hash_dir)):
fullname = join(hash_dir, wal_name)
if isdir(fullname):
_logger.warning(
'unexpected directory '
'rebuilding the wal database: %s',
fullname)
else:
if xlog.is_wal_file(fullname):
wal_count += 1
elif xlog.is_backup_file(fullname):
label_count += 1
else:
_logger.warning(
'unexpected file '
'rebuilding the wal database: %s',
fullname)
continue
wal_info = WalFileInfo.from_file(
fullname,
default_compression=default_compression)
fxlogdb_new.write(wal_info.to_xlogdb_line())
else:
# only history files are here
if xlog.is_history_file(fullname):
history_count += 1
wal_info = WalFileInfo.from_file(
fullname,
default_compression=default_compression)
fxlogdb_new.write(wal_info.to_xlogdb_line())
else:
_logger.warning(
'unexpected file '
'rebuilding the wal database: %s',
fullname)
os.fsync(fxlogdb_new.fileno())
shutil.move(xlogdb_new, fxlogdb.name)
fsync_dir(os.path.dirname(fxlogdb.name))
output.info('Done rebuilding xlogdb for server %s '
'(history: %s, backup_labels: %s, wal_file: %s)',
self.config.name, history_count, label_count, wal_count)
def get_latest_archived_wals_info(self):
"""
Return a dictionary of timelines associated with the
WalFileInfo of the last WAL file in the archive,
or None if the archive doesn't contain any WAL file.
:rtype: dict[str, WalFileInfo]|None
"""
from os.path import isdir, join
root = self.config.wals_directory
# If the WAL archive directory doesn't exists the archive is empty
if not isdir(root):
return None
# Traverse all the directory in the archive in reverse order,
# returning the first WAL file found
timelines = {}
for name in sorted(os.listdir(root), reverse=True):
fullname = join(root, name)
# All relevant files are in subdirectories, so
# we skip any non-directory entry
if isdir(fullname):
# Extract the timeline. If it is not valid, skip this directory
try:
timeline = name[0:8]
int(timeline, 16)
except ValueError:
continue
# If this timeline already has a file, skip this directory
if timeline in timelines:
continue
hash_dir = fullname
# Inspect contained files in reverse order
for wal_name in sorted(os.listdir(hash_dir), reverse=True):
fullname = join(hash_dir, wal_name)
# Return the first file that has the correct name
if not isdir(fullname) and xlog.is_wal_file(fullname):
timelines[timeline] = WalFileInfo.from_file(fullname)
break
# Return the timeline map or None if it is empty
return timelines or None
def remove_wal_before_backup(self, backup_info, timelines_to_protect=None):
"""
Remove WAL files which have been archived before the start of
the provided backup.
If no backup_info is provided delete all available WAL files
If timelines_to_protect list is passed, never remove a wal in one of
these timelines.
:param BackupInfo|None backup_info: the backup information structure
:param set timelines_to_protect: optional list of timelines
to protect
:return list: a list of removed WAL files
"""
removed = []
with self.server.xlogdb() as fxlogdb:
xlogdb_new = fxlogdb.name + ".new"
with open(xlogdb_new, 'w') as fxlogdb_new:
for line in fxlogdb:
wal_info = WalFileInfo.from_xlogdb_line(line)
if not xlog.is_any_xlog_file(wal_info.name):
output.error(
"invalid xlog segment name %r\n"
"HINT: Please run \"barman rebuild-xlogdb %s\" "
"to solve this issue",
wal_info.name, self.config.name)
continue
# Keeps the WAL segment if it is a history file
keep = xlog.is_history_file(wal_info.name)
# Keeps the WAL segment if its timeline is in
# `timelines_to_protect`
if timelines_to_protect:
tli, _, _ = xlog.decode_segment_name(wal_info.name)
keep |= tli in timelines_to_protect
# Keeps the WAL segment if it is a newer
# than the given backup (the first available)
if backup_info:
keep |= wal_info.name >= backup_info.begin_wal
# If the file has to be kept write it in the new xlogdb
# otherwise delete it and record it in the removed list
if keep:
fxlogdb_new.write(wal_info.to_xlogdb_line())
else:
self.delete_wal(wal_info)
removed.append(wal_info.name)
fxlogdb_new.flush()
os.fsync(fxlogdb_new.fileno())
shutil.move(xlogdb_new, fxlogdb.name)
fsync_dir(os.path.dirname(fxlogdb.name))
return removed
def validate_last_backup_maximum_age(self, last_backup_maximum_age):
"""
Evaluate the age of the last available backup in a catalogue.
If the last backup is older than the specified time interval (age),
the function returns False. If within the requested age interval,
the function returns True.
:param timedate.timedelta last_backup_maximum_age: time interval
representing the maximum allowed age for the last backup
in a server catalogue
:return tuple: a tuple containing the boolean result of the check and
auxiliary information about the last backup current age
"""
# Get the ID of the last available backup
backup_id = self.get_last_backup_id()
if backup_id:
# Get the backup object
backup = BackupInfo(self.server, backup_id=backup_id)
now = datetime.datetime.now(dateutil.tz.tzlocal())
# Evaluate the point of validity
validity_time = now - last_backup_maximum_age
# Pretty print of a time interval (age)
msg = human_readable_timedelta(now - backup.end_time)
# If the backup end time is older than the point of validity,
# return False, otherwise return true
if backup.end_time < validity_time:
return False, msg
else:
return True, msg
else:
# If no backup is available return false
return False, "No available backups"
def backup_fsync_and_set_sizes(self, backup_info):
"""
Fsync all files in a backup and set the actual size on disk
of a backup.
Also evaluate the deduplication ratio and the deduplicated size if
applicable.
:param barman.infofile.BackupInfo backup_info: the backup to update
"""
# Calculate the base backup size
self.executor.current_action = "calculating backup size"
_logger.debug(self.executor.current_action)
backup_size = 0
deduplicated_size = 0
backup_dest = backup_info.get_basebackup_directory()
for dir_path, _, file_names in os.walk(backup_dest):
# execute fsync() on the containing directory
fsync_dir(dir_path)
# execute fsync() on all the contained files
for filename in file_names:
file_path = os.path.join(dir_path, filename)
file_fd = os.open(file_path, os.O_RDONLY)
file_stat = os.fstat(file_fd)
backup_size += file_stat.st_size
# Excludes hard links from real backup size
if file_stat.st_nlink == 1:
deduplicated_size += file_stat.st_size
os.fsync(file_fd)
os.close(file_fd)
# Save size into BackupInfo object
backup_info.set_attribute('size', backup_size)
backup_info.set_attribute('deduplicated_size', deduplicated_size)
if backup_info.size > 0:
deduplication_ratio = 1 - (float(
backup_info.deduplicated_size) / backup_info.size)
else:
deduplication_ratio = 0
if self.config.reuse_backup == 'link':
output.info(
"Backup size: %s. Actual size on disk: %s"
" (-%s deduplication ratio)." % (
pretty_size(backup_info.size),
pretty_size(backup_info.deduplicated_size),
'{percent:.2%}'.format(percent=deduplication_ratio)
))
else:
output.info("Backup size: %s" %
pretty_size(backup_info.size))
barman-2.1/barman/wal_archiver.py 0000664 0001750 0001750 00000110546 13015546105 017501 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see
import collections
import datetime
import errno
import filecmp
import logging
import os
import shutil
from abc import ABCMeta, abstractmethod
from glob import glob
from distutils.version import LooseVersion as Version
from barman import output, xlog
from barman.command_wrappers import CommandFailedException, PgReceiveXlog
from barman.exceptions import (AbortedRetryHookScript, ArchiverFailure,
DuplicateWalFile, MatchingDuplicateWalFile)
from barman.hooks import HookScriptRunner, RetryHookScriptRunner
from barman.infofile import WalFileInfo
from barman.remote_status import RemoteStatusMixin
from barman.utils import fsync_dir, mkpath, with_metaclass
_logger = logging.getLogger(__name__)
class WalArchiverQueue(list):
def __init__(self, items, errors=None, skip=None, batch_size=0):
"""
A WalArchiverQueue is a list of WalFileInfo which has two extra
attribute list:
* errors: containing a list of unrecognized files
* skip: containing a list of skipped files.
It also stores batch run size information in case
it is requested by configuration, in order to limit the
number of WAL files that are processed in a single
run of the archive-wal command.
:param items: iterable from which initialize the list
:param batch_size: size of the current batch run (0=unlimited)
:param errors: an optional list of unrecognized files
:param skip: an optional list of skipped files
"""
super(WalArchiverQueue, self).__init__(items)
self.skip = []
self.errors = []
if skip is not None:
self.skip = skip
if errors is not None:
self.errors = errors
# Normalises batch run size
if batch_size > 0:
self.batch_size = batch_size
else:
self.batch_size = 0
@property
def size(self):
"""
Number of valid WAL segments waiting to be processed (in total)
:return int: total number of valid WAL files
"""
return len(self)
@property
def run_size(self):
"""
Number of valid WAL files to be processed in this run - takes
in consideration the batch size
:return int: number of valid WAL files for this batch run
"""
# In case a batch size has been explicitly specified
# (i.e. batch_size > 0), returns the minimum number between
# batch size and the queue size. Otherwise, simply
# returns the total queue size (unlimited batch size).
if self.batch_size > 0:
return min(self.size, self.batch_size)
return self.size
class WalArchiver(with_metaclass(ABCMeta, RemoteStatusMixin)):
"""
Base class for WAL archiver objects
"""
def __init__(self, backup_manager, name):
"""
Base class init method.
:param backup_manager: The backup manager
:param name: The name of this archiver
:return:
"""
self.backup_manager = backup_manager
self.server = backup_manager.server
self.config = backup_manager.config
self.name = name
super(WalArchiver, self).__init__()
def receive_wal(self, reset=False):
"""
Manage reception of WAL files. Does nothing by default.
Some archiver classes, like the StreamingWalArchiver, have a full
implementation.
:param bool reset: When set, resets the status of receive-wal
:raise ArchiverFailure: when something goes wrong
"""
def archive(self, fxlogdb, verbose=True):
"""
Archive WAL files, discarding duplicates or those that are not valid.
:param file fxlogdb: File object for xlogdb interactions
:param boolean verbose: Flag for verbose output
"""
compressor = self.backup_manager.compression_manager.get_compressor()
stamp = datetime.datetime.utcnow().strftime('%Y%m%dT%H%M%SZ')
processed = 0
header = "Processing xlog segments from %s for %s" % (
self.name, self.config.name)
# Get the next batch of WAL files to be processed
batch = self.get_next_batch()
# Analyse the batch and properly log the information
if batch.size:
if batch.size > batch.run_size:
# Batch mode enabled
_logger.info("Found %s xlog segments from %s for %s."
" Archive a batch of %s segments in this run.",
batch.size,
self.name,
self.config.name,
batch.run_size)
header += " (batch size: %s)" % batch.run_size
else:
# Single run mode (traditional)
_logger.info("Found %s xlog segments from %s for %s."
" Archive all segments in one run.",
batch.size,
self.name,
self.config.name)
else:
_logger.info("No xlog segments found from %s for %s.",
self.name,
self.config.name)
# Print the header (verbose mode)
if verbose:
output.info(header, log=False)
# Loop through all available WAL files
for wal_info in batch:
# Print the header (non verbose mode)
if not processed and not verbose:
output.info(header, log=False)
# Exit when archive batch size is reached
if processed >= batch.run_size:
_logger.debug("Batch size reached (%s) - "
"Exit %s process for %s",
batch.batch_size,
self.name,
self.config.name)
break
processed += 1
# Report to the user the WAL file we are archiving
output.info("\t%s", wal_info.name, log=False)
_logger.info("Archiving segment %s of %s from %s: %s/%s",
processed, batch.run_size, self.name,
self.config.name, wal_info.name)
# Archive the WAL file
try:
self.archive_wal(compressor, wal_info)
except MatchingDuplicateWalFile:
# We already have this file. Simply unlink the file.
os.unlink(wal_info.orig_filename)
continue
except DuplicateWalFile:
output.info("\tError: %s is already present in server %s. "
"File moved to errors directory.",
wal_info.name,
self.config.name)
error_dst = os.path.join(
self.config.errors_directory,
"%s.%s.duplicate" % (wal_info.name,
stamp))
# TODO: cover corner case of duplication (unlikely,
# but theoretically possible)
shutil.move(wal_info.orig_filename, error_dst)
continue
except AbortedRetryHookScript as e:
_logger.warning("Archiving of %s/%s aborted by "
"pre_archive_retry_script."
"Reason: %s" % (self.config.name,
wal_info.name,
e))
return
# Updates the information of the WAL archive with
# the latest segments
fxlogdb.write(wal_info.to_xlogdb_line())
# flush and fsync for every line
fxlogdb.flush()
os.fsync(fxlogdb.fileno())
if processed:
_logger.debug("Archived %s out of %s xlog segments from %s for %s",
processed, batch.size, self.name, self.config.name)
elif verbose:
output.info("\tno file found", log=False)
if batch.errors:
output.info("Some unknown objects have been found while "
"processing xlog segments for %s. "
"Objects moved to errors directory:",
self.config.name,
log=False)
# Log unexpected files
_logger.warning("Archiver is about to move %s unexpected file(s) "
"to errors directory for %s from %s",
len(batch.errors),
self.config.name,
self.name)
for error in batch.errors:
basename = os.path.basename(error)
output.info("\t%s", basename, log=False)
# Print informative log line.
_logger.warning("Moving unexpected file for %s from %s: %s",
self.config.name, self.name, basename)
error_dst = os.path.join(
self.config.errors_directory,
"%s.%s.unknown" % (basename, stamp))
try:
shutil.move(error, error_dst)
except IOError as e:
if e.errno == errno.ENOENT:
_logger.warning('%s not found' % error)
def archive_wal(self, compressor, wal_info):
"""
Archive a WAL segment and update the wal_info object
:param compressor: the compressor for the file (if any)
:param WalFileInfo wal_info: the WAL file is being processed
"""
src_file = wal_info.orig_filename
src_dir = os.path.dirname(src_file)
dst_file = wal_info.fullpath(self.server)
tmp_file = dst_file + '.tmp'
dst_dir = os.path.dirname(dst_file)
error = None
try:
# Run the pre_archive_script if present.
script = HookScriptRunner(self.backup_manager,
'archive_script', 'pre')
script.env_from_wal_info(wal_info, src_file)
script.run()
# Run the pre_archive_retry_script if present.
retry_script = RetryHookScriptRunner(self.backup_manager,
'archive_retry_script',
'pre')
retry_script.env_from_wal_info(wal_info, src_file)
retry_script.run()
# Check if destination already exists
if os.path.exists(dst_file):
src_uncompressed = src_file
dst_uncompressed = dst_file
dst_info = WalFileInfo.from_file(dst_file)
try:
comp_manager = self.backup_manager.compression_manager
if dst_info.compression is not None:
dst_uncompressed = dst_file + '.uncompressed'
comp_manager.get_compressor(
compression=dst_info.compression).decompress(
dst_file, dst_uncompressed)
if wal_info.compression:
src_uncompressed = src_file + '.uncompressed'
comp_manager.get_compressor(
compression=wal_info.compression).decompress(
src_file, src_uncompressed)
# Directly compare files.
# When the files are identical
# raise a MatchingDuplicateWalFile exception,
# otherwise raise a DuplicateWalFile exception.
if filecmp.cmp(dst_uncompressed, src_uncompressed):
raise MatchingDuplicateWalFile(wal_info)
else:
raise DuplicateWalFile(wal_info)
finally:
if src_uncompressed != src_file:
os.unlink(src_uncompressed)
if dst_uncompressed != dst_file:
os.unlink(dst_uncompressed)
mkpath(dst_dir)
# Compress the file only if not already compressed
if compressor and not wal_info.compression:
compressor.compress(src_file, tmp_file)
shutil.copystat(src_file, tmp_file)
os.rename(tmp_file, dst_file)
os.unlink(src_file)
# Update wal_info
stat = os.stat(dst_file)
wal_info.size = stat.st_size
wal_info.compression = compressor.compression
else:
# Try to atomically rename the file. If successful,
# the renaming will be an atomic operation
# (this is a POSIX requirement).
try:
os.rename(src_file, dst_file)
except OSError:
# Source and destination are probably on different
# filesystems
shutil.copy2(src_file, tmp_file)
os.rename(tmp_file, dst_file)
os.unlink(src_file)
# At this point the original file has been removed
wal_info.orig_filename = None
# Execute fsync() on the archived WAL file
file_fd = os.open(dst_file, os.O_RDONLY)
os.fsync(file_fd)
os.close(file_fd)
# Execute fsync() on the archived WAL containing directory
fsync_dir(dst_dir)
# Execute fsync() also on the incoming directory
fsync_dir(src_dir)
except Exception as e:
# In case of failure save the exception for the post scripts
error = e
raise
# Ensure the execution of the post_archive_retry_script and
# the post_archive_script
finally:
# Run the post_archive_retry_script if present.
try:
retry_script = RetryHookScriptRunner(self,
'archive_retry_script',
'post')
retry_script.env_from_wal_info(wal_info, dst_file, error)
retry_script.run()
except AbortedRetryHookScript as e:
# Ignore the ABORT_STOP as it is a post-hook operation
_logger.warning("Ignoring stop request after receiving "
"abort (exit code %d) from post-archive "
"retry hook script: %s",
e.hook.exit_status, e.hook.script)
# Run the post_archive_script if present.
script = HookScriptRunner(self, 'archive_script', 'post', error)
script.env_from_wal_info(wal_info, dst_file)
script.run()
@abstractmethod
def get_next_batch(self):
"""
Return a WalArchiverQueue containing the WAL files to be archived.
:rtype: WalArchiverQueue
"""
@abstractmethod
def check(self, check_strategy):
"""
Perform specific checks for the archiver - invoked
by server.check_postgres
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
@abstractmethod
def status(self):
"""
Set additional status info - invoked by Server.status()
"""
@staticmethod
def summarise_error_files(error_files):
"""
Summarise a error files list
:param list[str] error_files: Error files list to summarise
:return str: A summary, None if there are no error files
"""
if not error_files:
return None
# The default value for this dictionary will be 0
counters = collections.defaultdict(int)
# Count the file types
for name in error_files:
if name.endswith(".error"):
counters['not relevant'] += 1
elif name.endswith(".duplicate"):
counters['duplicates'] += 1
elif name.endswith(".unknown"):
counters['unknown'] += 1
else:
counters['unknown failure'] += 1
# Return a summary list of the form: "item a: 2, item b: 5"
return ', '.join("%s: %s" % entry for entry in counters.items())
class FileWalArchiver(WalArchiver):
"""
Manager of file-based WAL archiving operations (aka 'log shipping').
"""
def __init__(self, backup_manager):
super(FileWalArchiver, self).__init__(backup_manager, 'file archival')
def fetch_remote_status(self):
"""
Returns the status of the FileWalArchiver.
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
result = dict.fromkeys(
['archive_mode', 'archive_command'], None)
postgres = self.server.postgres
# If Postgres is not available we cannot detect anything
if not postgres:
return result
# Query the database for 'archive_mode' and 'archive_command'
result['archive_mode'] = postgres.get_setting('archive_mode')
result['archive_command'] = postgres.get_setting('archive_command')
# Add pg_stat_archiver statistics if the view is supported
pg_stat_archiver = postgres.get_archiver_stats()
if pg_stat_archiver is not None:
result.update(pg_stat_archiver)
return result
def get_next_batch(self):
"""
Returns the next batch of WAL files that have been archived through
a PostgreSQL's 'archive_command' (in the 'incoming' directory)
:return: WalArchiverQueue: list of WAL files
"""
# Get the batch size from configuration (0 = unlimited)
batch_size = self.config.archiver_batch_size
# List and sort all files in the incoming directory
file_names = glob(os.path.join(
self.config.incoming_wals_directory, '*'))
file_names.sort()
# Process anything that looks like a valid WAL file. Anything
# else is treated like an error/anomaly
files = []
errors = []
for file_name in file_names:
if xlog.is_any_xlog_file(file_name) and os.path.isfile(file_name):
files.append(file_name)
else:
errors.append(file_name)
# Build the list of WalFileInfo
wal_files = [WalFileInfo.from_file(f) for f in files]
return WalArchiverQueue(wal_files,
batch_size=batch_size,
errors=errors)
def check(self, check_strategy):
"""
Perform additional checks for FileWalArchiver - invoked
by server.check_postgres
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
remote_status = self.get_remote_status()
# If archive_mode is None, there are issues connecting to PostgreSQL
if remote_status['archive_mode'] is None:
return
# Check archive_mode parameter: must be on
if remote_status['archive_mode'] in ('on', 'always'):
check_strategy.result(self.config.name, 'archive_mode', True)
else:
msg = "please set it to 'on'"
if self.server.postgres.server_version >= 90500:
msg += " or 'always'"
check_strategy.result(self.config.name, 'archive_mode', False, msg)
if remote_status['archive_command'] and \
remote_status['archive_command'] != '(disabled)':
check_strategy.result(self.config.name, 'archive_command',
True)
# Report if the archiving process works without issues.
# Skip if the archive_command check fails
# It can be None if PostgreSQL is older than 9.4
if remote_status.get('is_archiving') is not None:
check_strategy.result(
self.config.name, 'continuous archiving',
remote_status['is_archiving'])
else:
check_strategy.result(
self.config.name, 'archive_command', False,
'please set it accordingly to documentation')
def status(self):
"""
Set additional status info - invoked by Server.status()
"""
remote_status = self.get_remote_status()
# If archive_mode is None, there are issues connecting to PostgreSQL
if remote_status['archive_mode'] is None:
return
output.result('status', self.config.name,
"archive_command",
"PostgreSQL 'archive_command' setting",
remote_status['archive_command'] or
"FAILED (please set it accordingly to documentation)")
last_wal = remote_status.get('last_archived_wal')
# If PostgreSQL is >= 9.4 we have the last_archived_time
if last_wal and remote_status.get('last_archived_time'):
last_wal += ", at %s" % (
remote_status['last_archived_time'].ctime())
output.result('status', self.config.name,
"last_archived_wal",
"Last archived WAL",
last_wal or "No WAL segment shipped yet")
# Set output for WAL archive failures (PostgreSQL >= 9.4)
if remote_status.get('failed_count') is not None:
remote_fail = str(remote_status['failed_count'])
if int(remote_status['failed_count']) > 0:
remote_fail += " (%s at %s)" % (
remote_status['last_failed_wal'],
remote_status['last_failed_time'].ctime())
output.result('status', self.config.name, 'failed_count',
'Failures of WAL archiver', remote_fail)
# Add hourly archive rate if available (PostgreSQL >= 9.4) and > 0
if remote_status.get('current_archived_wals_per_second'):
output.result(
'status', self.config.name,
'server_archived_wals_per_hour',
'Server WAL archiving rate', '%0.2f/hour' % (
3600 * remote_status['current_archived_wals_per_second']))
class StreamingWalArchiver(WalArchiver):
"""
Object used for the management of streaming WAL archive operation.
"""
def __init__(self, backup_manager):
super(StreamingWalArchiver, self).__init__(backup_manager, 'streaming')
def fetch_remote_status(self):
"""
Execute checks for replication-based wal archiving
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
remote_status = dict.fromkeys(
('pg_receivexlog_compatible',
'pg_receivexlog_installed',
'pg_receivexlog_path',
'pg_receivexlog_supports_slots',
'pg_receivexlog_synchronous',
'pg_receivexlog_version'),
None)
# Test pg_receivexlog existence
version_info = PgReceiveXlog.get_version_info(
self.server.path)
if version_info['full_path']:
remote_status["pg_receivexlog_installed"] = True
remote_status["pg_receivexlog_path"] = version_info['full_path']
remote_status["pg_receivexlog_version"] = (
version_info['full_version'])
pgreceivexlog_version = version_info['major_version']
else:
remote_status["pg_receivexlog_installed"] = False
return remote_status
# Retrieve the PostgreSQL version
pg_version = None
if self.server.streaming is not None:
pg_version = self.server.streaming.server_major_version
# If one of the version is unknown we cannot compare them
if pgreceivexlog_version is None or pg_version is None:
return remote_status
# pg_version is not None so transform into a Version object
# for easier comparison between versions
pg_version = Version(pg_version)
# Set conservative default values (False) for modern features
remote_status["pg_receivexlog_compatible"] = False
remote_status['pg_receivexlog_supports_slots'] = False
remote_status["pg_receivexlog_synchronous"] = False
# pg_receivexlog 9.2 is compatible only with PostgreSQL 9.2.
if "9.2" == pg_version == pgreceivexlog_version:
remote_status["pg_receivexlog_compatible"] = True
# other versions are compatible with lesser versions of PostgreSQL
# WARNING: The development versions of `pg_receivexlog` are considered
# higher than the stable versions here, but this is not an issue
# because it accepts everything that is less than
# the `pg_receivexlog` version(e.g. '9.6' is less than '9.6devel')
elif "9.2" < pg_version <= pgreceivexlog_version:
# At least PostgreSQL 9.3 is required here
remote_status["pg_receivexlog_compatible"] = True
# replication slots are supported starting from version 9.4
if "9.4" <= pg_version <= pgreceivexlog_version:
remote_status['pg_receivexlog_supports_slots'] = True
# Synchronous WAL streaming requires replication slots
# and pg_receivexlog >= 9.5
if "9.4" <= pg_version and "9.5" <= pgreceivexlog_version:
remote_status["pg_receivexlog_synchronous"] = (
self._is_synchronous())
return remote_status
def receive_wal(self, reset=False):
"""
Creates a PgReceiveXlog object and issues the pg_receivexlog command
for a specific server
:param bool reset: When set reset the status of receive-wal
:raise ArchiverFailure: when something goes wrong
"""
# Ensure the presence of the destination directory
mkpath(self.config.streaming_wals_directory)
# Check if is a reset request
if reset:
self._reset_streaming_status()
return
# Execute basic sanity checks on PostgreSQL connection
streaming_status = self.server.streaming.get_remote_status()
if streaming_status["streaming_supported"] is None:
raise ArchiverFailure(
'failed opening the PostgreSQL streaming connection')
elif not streaming_status["streaming_supported"]:
raise ArchiverFailure(
'PostgreSQL version too old (%s < 9.2)' %
self.server.streaming.server_txt_version)
# Execute basic sanity checks on pg_receivexlog
remote_status = self.get_remote_status()
if not remote_status["pg_receivexlog_installed"]:
raise ArchiverFailure(
'pg_receivexlog not present in $PATH')
if not remote_status['pg_receivexlog_compatible']:
raise ArchiverFailure(
'pg_receivexlog version not compatible with '
'PostgreSQL server version')
# Execute sanity check on replication slot usage
if self.config.slot_name:
# Check if slots are supported
if not remote_status['pg_receivexlog_supports_slots']:
raise ArchiverFailure(
'Physical replication slot not supported by %s '
'(9.4 or higher is required)' %
self.server.streaming.server_txt_version)
# Check if the required slot exists
postgres_status = self.server.postgres.get_remote_status()
if postgres_status['replication_slot'] is None:
raise ArchiverFailure(
"replication slot '%s' doesn't exist. "
"Please execute "
"'barman receive-wal --create-slot %s'" %
(self.config.slot_name, self.config.name))
# Check if the required slot is available
if postgres_status['replication_slot'].active:
raise ArchiverFailure(
"replication slot '%s' is already in use" %
(self.config.slot_name,))
# Make sure we are not wasting precious PostgreSQL resources
self.server.close()
_logger.info('Activating WAL archiving through streaming protocol')
try:
output_handler = PgReceiveXlog.make_output_handler(
self.config.name + ': ')
receive = PgReceiveXlog(
connection=self.server.streaming,
destination=self.config.streaming_wals_directory,
command=remote_status['pg_receivexlog_path'],
version=remote_status['pg_receivexlog_version'],
app_name=self.config.streaming_archiver_name,
path=self.server.path,
slot_name=self.config.slot_name,
synchronous=remote_status['pg_receivexlog_synchronous'],
out_handler=output_handler,
err_handler=output_handler
)
# Finally execute the pg_receivexlog process
receive.execute()
except CommandFailedException as e:
# Retrieve the return code from the exception
ret_code = e.args[0]['ret']
if ret_code < 0:
# If the return code is negative, then pg_receivexlog
# was terminated by a signal
msg = "pg_receivexlog terminated by signal: %s" \
% abs(ret_code)
else:
# Otherwise terminated with an error
msg = "pg_receivexlog terminated with error code: %s"\
% ret_code
raise ArchiverFailure(msg)
except KeyboardInterrupt:
# This is a normal termination, so there is nothing to do beside
# informing the user.
output.info('SIGINT received. Terminate gracefully.')
def _reset_streaming_status(self):
"""
Reset the status of receive-wal removing any .partial files
"""
output.info("Resetting receive-wal directory status")
partial_files = glob(os.path.join(
self.config.streaming_wals_directory, '*.partial'))
for partial in partial_files:
output.info("Removing status file %s" % partial)
os.unlink(partial)
def get_next_batch(self):
"""
Returns the next batch of WAL files that have been archived via
streaming replication (in the 'streaming' directory)
This method always leaves one file in the "streaming" directory,
because the 'pg_receivexlog' process needs at least one file to
detect the current streaming position after a restart.
:return: WalArchiverQueue: list of WAL files
"""
# Get the batch size from configuration (0 = unlimited)
batch_size = self.config.streaming_archiver_batch_size
# List and sort all files in the incoming directory
file_names = glob(os.path.join(
self.config.streaming_wals_directory, '*'))
file_names.sort()
# Process anything that looks like a valid WAL file,
# including partial ones and history files.
# Anything else is treated like an error/anomaly
files = []
skip = []
errors = []
for file_name in file_names:
if not os.path.isfile(file_name):
errors.append(file_name)
elif xlog.is_partial_file(file_name):
skip.append(file_name)
elif xlog.is_any_xlog_file(file_name):
files.append(file_name)
else:
errors.append(file_name)
# In case of more than a partial file, keep the last
# and treat the rest as errors
if len(skip) > 1:
errors.extend(skip[:-1])
_logger.warning('Multiple partial files found for server %s: %s' %
(self.config.name,
", ".join([os.path.basename(f) for f in errors])))
skip = skip[-1:]
# Keep the last full WAL file in case no partial file is present
elif len(skip) == 0 and files:
skip.append(files.pop())
# Build the list of WalFileInfo
wal_files = [WalFileInfo.from_file(f, compression=None) for f in files]
return WalArchiverQueue(wal_files,
batch_size=batch_size,
errors=errors, skip=skip)
def check(self, check_strategy):
"""
Perform additional checks for FileWalArchiver - invoked
by server.check_postgres
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
# Check the version of pg_receivexlog
remote_status = self.get_remote_status()
check_strategy.result(
self.config.name, 'pg_receivexlog',
remote_status['pg_receivexlog_installed'])
hint = None
if not remote_status['pg_receivexlog_compatible']:
pg_version = 'Unknown'
if self.server.streaming is not None:
pg_version = self.server.streaming.server_txt_version
hint = "PostgreSQL version: %s, pg_receivexlog version: %s" % (
pg_version,
remote_status['pg_receivexlog_version']
)
check_strategy.result(
self.config.name, 'pg_receivexlog compatible',
remote_status['pg_receivexlog_compatible'], hint=hint)
# Check if pg_receivexlog is running, by retrieving a list
# of running 'receive-wal' processes from the process manager.
receiver_list = self.server.process_manager.list('receive-wal')
# If there's at least one 'receive-wal' process running for this
# server, the test is passed
if receiver_list:
check_strategy.result(
self.config.name, 'receive-wal running', True)
else:
check_strategy.result(
self.config.name, 'receive-wal running', False,
'See the Barman log file for more details')
def _is_synchronous(self):
"""
Check if receive-wal process is eligible for synchronous replication
The receive-wal process is eligible for synchronous replication
if `synchronous_standby_names` is configured and contains
the value of `streaming_archiver_name`
:rtype: bool
"""
# Nothing to do if postgres connection is not working
postgres = self.server.postgres
if postgres is None or postgres.server_txt_version is None:
return None
# Check if synchronous WAL streaming can be enabled
# by peeking 'synchronous_standby_names'
postgres_status = postgres.get_remote_status()
syncnames = postgres_status['synchronous_standby_names']
_logger.debug("Look for '%s' in "
"'synchronous_standby_names': %s",
self.config.streaming_archiver_name, syncnames)
# The receive-wal process is eligible for synchronous replication
# if `synchronous_standby_names` is configured and contains
# the value of `streaming_archiver_name`
synchronous = (syncnames and
self.config.streaming_archiver_name in syncnames)
_logger.info('Synchronous WAL streaming for %s: %s',
self.config.streaming_archiver_name,
synchronous)
return synchronous
def status(self):
"""
Set additional status info - invoked by Server.status()
"""
# TODO: Add status information for WAL streaming
barman-2.1/barman/command_wrappers.py 0000664 0001750 0001750 00000105036 13026732452 020376 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module contains a wrapper for shell commands
"""
from __future__ import print_function
import errno
import inspect
import logging
import os
import select
import signal
import subprocess
import sys
import time
from distutils.version import LooseVersion as Version
import barman.utils
from barman.exceptions import CommandFailedException, CommandMaxRetryExceeded
_logger = logging.getLogger(__name__)
class StreamLineProcessor(object):
"""
Class deputed to reading lines from a file object, using a buffered read.
NOTE: This class never call os.read() twice in a row. And is designed to
work with the select.select() method.
"""
def __init__(self, fobject, handler):
"""
:param file fobject: The file that is being read
:param callable handler: The function (taking only one unicode string
argument) which will be called for every line
"""
self._file = fobject
self._handler = handler
self._buf = ''
def fileno(self):
"""
Method used by select.select() to get the underlying file descriptor.
:rtype: the underlying file descriptor
"""
return self._file.fileno()
def process(self):
"""
Read the ready data from the stream and for each line found invoke the
handler.
:return bool: True when End Of File has been reached
"""
data = os.read(self._file.fileno(), 4096)
# If nothing has been read, we reached the EOF
if not data:
self._file.close()
# Handle the last line (always incomplete, maybe empty)
self._handler(self._buf)
return True
self._buf += data.decode('utf-8')
# If no '\n' is present, we just read a part of a very long line.
# Nothing to do at the moment.
if '\n' not in self._buf:
return False
tmp = self._buf.split('\n')
# Leave the remainder in self._buf
self._buf = tmp[-1]
# Call the handler for each complete line.
lines = tmp[:-1]
for line in lines:
self._handler(line)
return False
class Command(object):
"""
Wrapper for a system command
"""
def __init__(self, cmd, args=None, env_append=None, path=None, shell=False,
check=False, allowed_retval=(0,),
close_fds=True, out_handler=None, err_handler=None,
retry_times=0, retry_sleep=0, retry_handler=None):
"""
If the `args` argument is specified the arguments will be always added
to the ones eventually passed with the actual invocation.
If the `env_append` argument is present its content will be appended to
the environment of every invocation.
The subprocess output and error stream will be processed through
the output and error handler, respectively defined through the
`out_handler` and `err_handler` arguments. If not provided every line
will be sent to teh log respectively at INFO and WARNING level.
The `out_handler` and the `err_handler` functions will be invoked with
one single argument, which is a string containing the line that is
being processed.
If the `close_fds` argument is True, all file descriptors
except 0, 1 and 2 will be closed before the child process is executed.
If the `check` argument is True, the exit code will be checked
against the `allowed_retval` list, raising a CommandFailedException if
not in the list.
If `retry_times` is greater than 0, when the execution of a command
terminates with an error, it will be retried for
a maximum of `retry_times` times, waiting for `retry_sleep` seconds
between every attempt.
Everytime a command is retried the `retry_handler` is executed
before running the command again. The retry_handler must be a callable
that accepts the following fields:
* the Command object
* the arguments list
* the keyword arguments dictionary
* the number of the failed attempt
* the exception containing the error
An example of such a function is:
> def retry_handler(command, args, kwargs, attempt, exc):
> print("Failed command!")
Some of the keyword arguments can be specified both in the class
constructor and during the method call. If specified in both places,
the method arguments will take the precedence over
the constructor arguments.
:param str cmd: The command to exexute
:param list[str]|None args: List of additional arguments to append
:param dict[str.str]|None env_append: additional environment variables
:param str path: PATH to be used while searching for `cmd`
:param bool shell: If true, use the shell instead of an "execve" call
:param bool check: Raise a CommandFailedException if the exit code
is not present in `allowed_retval`
:param list[int] allowed_retval: List of exit codes considered as a
successful termination.
:param bool close_fds: If set, close all the extra file descriptors
:param callable out_handler: handler for lines sent on stdout
:param callable err_handler: handler for lines sent on stderr
:param int retry_times: number of allowed retry attempts
:param int retry_sleep: wait seconds between every retry
:param callable retry_handler: handler invoked during a command retry
"""
self.pipe = None
self.cmd = cmd
self.args = args if args is not None else []
self.shell = shell
self.close_fds = close_fds
self.check = check
self.allowed_retval = allowed_retval
self.retry_times = retry_times
self.retry_sleep = retry_sleep
self.retry_handler = retry_handler
self.path = path
self.ret = None
self.out = None
self.err = None
# If env_append has been provided use it or replace with an empty dict
env_append = env_append or {}
# If path has been provided, replace it in the environment
if path:
env_append['PATH'] = path
# Find the absolute path to the command to execute
if not self.shell:
full_path = barman.utils.which(self.cmd, self.path)
if not full_path:
raise CommandFailedException(
'%s not in PATH' % self.cmd)
self.cmd = full_path
# If env_append contains anything, build an env dict to be used during
# subprocess call, otherwise set it to None and let the subprocesses
# inherit the parent environment
if env_append:
self.env = os.environ.copy()
self.env.update(env_append)
else:
self.env = None
# If an output handler has been provided use it, otherwise log the
# stdout as INFO
if out_handler:
self.out_handler = out_handler
else:
self.out_handler = self.make_logging_handler(logging.INFO)
# If an error handler has been provided use it, otherwise log the
# stderr as WARNING
if err_handler:
self.err_handler = err_handler
else:
self.err_handler = self.make_logging_handler(logging.WARNING)
@staticmethod
def _restore_sigpipe():
"""restore default signal handler (http://bugs.python.org/issue1652)"""
signal.signal(signal.SIGPIPE, signal.SIG_DFL) # pragma: no cover
@staticmethod
def _cmd_quote(cmd, args):
"""
Quote all cmd's arguments.
This is needed to avoid command string breaking.
WARNING: this function does not protect against injection.
"""
if args is not None and len(args) > 0:
cmd = "%s '%s'" % (cmd, "' '".join(args))
return cmd
def __call__(self, *args, **kwargs):
"""
Run the command and return the exit code.
The output and error strings are not returned, but they can be accessed
as attributes of the Command object, as well as the exit code.
If `stdin` argument is specified, its content will be passed to the
executed command through the standard input descriptor.
If the `close_fds` argument is True, all file descriptors
except 0, 1 and 2 will be closed before the child process is executed.
If the `check` argument is True, the exit code will be checked
against the `allowed_retval` list, raising a CommandFailedException if
not in the list.
Every keyword argument can be specified both in the class constructor
and during the method call. If specified in both places,
the method arguments will take the precedence over
the constructor arguments.
:rtype: int
:raise: CommandFailedException
:raise: CommandMaxRetryExceeded
"""
self.get_output(*args, **kwargs)
return self.ret
def get_output(self, *args, **kwargs):
"""
Run the command and return the output and the error as a tuple.
The return code is not returned, but it can be accessed as an attribute
of the Command object, as well as the output and the error strings.
If `stdin` argument is specified, its content will be passed to the
executed command through the standard input descriptor.
If the `close_fds` argument is True, all file descriptors
except 0, 1 and 2 will be closed before the child process is executed.
If the `check` argument is True, the exit code will be checked
against the `allowed_retval` list, raising a CommandFailedException if
not in the list.
Every keyword argument can be specified both in the class constructor
and during the method call. If specified in both places,
the method arguments will take the precedence over
the constructor arguments.
:rtype: tuple[str, str]
:raise: CommandFailedException
:raise: CommandMaxRetryExceeded
"""
attempt = 0
while True:
try:
return self._get_output_once(*args, **kwargs)
except CommandFailedException as exc:
# Try again if retry number is lower than the retry limit
if attempt < self.retry_times:
# If a retry_handler is defined, invoke it passing the
# Command instance and the exception
if self.retry_handler:
self.retry_handler(self, args, kwargs, attempt, exc)
# Sleep for configured time, then try again
time.sleep(self.retry_sleep)
attempt += 1
else:
if attempt == 0:
# No retry requested by the user
# Raise the original exception
raise
else:
# If the max number of attempts is reached and
# there is still an error, exit raising
# a CommandMaxRetryExceeded exception and wrap the
# original one
raise CommandMaxRetryExceeded(exc)
def _get_output_once(self, *args, **kwargs):
"""
Run the command and return the output and the error as a tuple.
The return code is not returned, but it can be accessed as an attribute
of the Command object, as well as the output and the error strings.
If `stdin` argument is specified, its content will be passed to the
executed command through the standard input descriptor.
If the `close_fds` argument is True, all file descriptors
except 0, 1 and 2 will be closed before the child process is executed.
If the `check` argument is True, the exit code will be checked
against the `allowed_retval` list, raising a CommandFailedException if
not in the list.
Every keyword argument can be specified both in the class constructor
and during the method call. If specified in both places,
the method arguments will take the precedence over
the constructor arguments.
:rtype: tuple[str, str]
:raises: CommandFailedException
"""
out = []
err = []
# If check is true, it must be handled here
check = kwargs.pop('check', self.check)
allowed_retval = kwargs.pop('allowed_retval', self.allowed_retval)
self.execute(out_handler=out.append, err_handler=err.append,
check=False, *args, **kwargs)
self.out = '\n'.join(out)
self.err = '\n'.join(err)
_logger.debug("Command stdout: %s", self.out)
_logger.debug("Command stderr: %s", self.err)
# Raise if check and the return code is not in the allowed list
if check:
self.check_return_value(allowed_retval)
return self.out, self.err
def check_return_value(self, allowed_retval):
"""
Check the current return code and raise CommandFailedException when
it's not in the allowed_retval list
:param list[int] allowed_retval: list of return values considered
success
:raises: CommandFailedException
"""
if self.ret not in allowed_retval:
raise CommandFailedException(dict(
ret=self.ret, out=self.out, err=self.err))
def execute(self, *args, **kwargs):
"""
Execute the command and pass the output to the configured handlers
If `stdin` argument is specified, its content will be passed to the
executed command through the standard input descriptor.
The subprocess output and error stream will be processed through
the output and error handler, respectively defined through the
`out_handler` and `err_handler` arguments. If not provided every line
will be sent to teh log respectively at INFO and WARNING level.
If the `close_fds` argument is True, all file descriptors
except 0, 1 and 2 will be closed before the child process is executed.
If the `check` argument is True, the exit code will be checked
against the `allowed_retval` list, raising a CommandFailedException if
not in the list.
Every keyword argument can be specified both in the class constructor
and during the method call. If specified in both places,
the method arguments will take the precedence over
the constructor arguments.
:rtype: int
:raise: CommandFailedException
"""
# Check keyword arguments
stdin = kwargs.pop('stdin', None)
check = kwargs.pop('check', self.check)
allowed_retval = kwargs.pop('allowed_retval', self.allowed_retval)
close_fds = kwargs.pop('close_fds', self.close_fds)
out_handler = kwargs.pop('out_handler', self.out_handler)
err_handler = kwargs.pop('err_handler', self.err_handler)
if len(kwargs):
raise TypeError('%s() got an unexpected keyword argument %r' %
(inspect.stack()[1][3], kwargs.popitem()[0]))
# Reset status
self.ret = None
self.out = None
self.err = None
# Create the subprocess and save it in the current object to be usable
# by signal handlers
pipe = self._build_pipe(args, close_fds)
self.pipe = pipe
# Send the provided input and close the stdin descriptor
if stdin:
pipe.stdin.write(stdin)
pipe.stdin.close()
# Prepare the list of processors
processors = [
StreamLineProcessor(
pipe.stdout, out_handler),
StreamLineProcessor(
pipe.stderr, err_handler)]
# Read the streams until the subprocess exits
self.pipe_processor_loop(processors)
# Reap the zombie and read the exit code
pipe.wait()
self.ret = pipe.returncode
# Remove the closed pipe from the object
self.pipe = None
_logger.debug("Command return code: %s", self.ret)
# Raise if check and the return code is not in the allowed list
if check:
self.check_return_value(allowed_retval)
return self.ret
def _build_pipe(self, args, close_fds):
"""
Build the Pipe object used by the Command
The resulting command will be composed by:
self.cmd + self.args + args
:param args: extra arguments for the subprocess
:param close_fds: if True all file descriptors except 0, 1 and 2
will be closed before the child process is executed.
:rtype: subprocess.Popen
"""
# Append the argument provided to this method ot the base argument list
args = self.args + list(args)
# If shell is True, properly quote the command
if self.shell:
cmd = self._cmd_quote(self.cmd, args)
else:
cmd = [self.cmd] + args
# Log the command we are about to execute
_logger.debug("Command: %r", cmd)
return subprocess.Popen(cmd, shell=self.shell, env=self.env,
stdin=subprocess.PIPE,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
preexec_fn=self._restore_sigpipe,
close_fds=close_fds)
@staticmethod
def pipe_processor_loop(processors):
"""
Process the output received through the pipe until all the provided
StreamLineProcessor reach the EOF.
:param list[StreamLineProcessor] processors: a list of
StreamLineProcessor
"""
# Loop until all the streams reaches the EOF
while processors:
try:
ready = select.select(processors, [], [])[0]
except select.error as e:
# If the select call has been interrupted by a signal
# just retry
if e.args[0] == errno.EINTR:
continue
raise
# For each ready StreamLineProcessor invoke the process() method
for stream in ready:
eof = stream.process()
# Got EOF on this stream
if eof:
# Remove the stream from the list of valid processors
processors.remove(stream)
@classmethod
def make_logging_handler(cls, level, prefix=None):
"""
Build a handler function that logs every line it receives.
The resulting function logs its input at the specified level
with an optional prefix.
:param level: The log level to use
:param prefix: An optional prefix to prepend to the line
:return: handler function
"""
class_logger = logging.getLogger(cls.__name__)
def handler(line):
if line:
if prefix:
class_logger.log(level, "%s%s", prefix, line)
else:
class_logger.log(level, "%s", line)
return handler
@staticmethod
def make_output_handler(prefix=None):
"""
Build a handler function which prints every line it receives.
The resulting function prints (and log it at INFO level) its input
with an optional prefix.
:param prefix: An optional prefix to prepend to the line
:return: handler function
"""
# Import the output module inside the function to avoid circular
# dependency
from barman import output
def handler(line):
if line:
if prefix:
output.info("%s%s", prefix, line)
else:
output.info("%s", line)
return handler
def enable_signal_forwarding(self, signal_id):
"""
Enable signal forwarding to the subprocess for a specified signal_id
:param signal_id: The signal id to be forwarded
"""
# Get the current signal handler
old_handler = signal.getsignal(signal_id)
def _handler(sig, frame):
"""
This signal handler forward the signal to the subprocess then
execute the original handler.
"""
# Forward the signal to the subprocess
if self.pipe:
self.pipe.send_signal(signal_id)
# If the old handler is callable
if callable(old_handler):
old_handler(sig, frame)
# If we have got a SIGTERM, we must exit
elif old_handler == signal.SIG_DFL and signal_id == signal.SIGTERM:
sys.exit(128 + signal_id)
# Set the signal handler
signal.signal(signal_id, _handler)
class Rsync(Command):
"""
This class is a wrapper for the rsync system command,
which is used vastly by barman
"""
def __init__(self, rsync='rsync', args=None, ssh=None, ssh_options=None,
bwlimit=None, exclude=None, exclude_and_protect=None,
include=None, network_compression=None, path=None, **kwargs):
"""
:param str rsync: rsync executable name
:param list[str]|None args: List of additional argument to aways append
:param str ssh: the ssh executable to be used when building
the `-e` argument
:param list[str] ssh_options: the ssh options to be used when building
the `-e` argument
:param str bwlimit: optional bandwidth limit
:param list[str] exclude: list of file to be excluded from the copy
:param list[str] exclude_and_protect: list of file to be excluded from
the copy, preserving the destination if exists
:param list[str] include: list of files to be included in the copy
even if excluded.
:param bool network_compression: enable the network compression
:param str path: PATH to be used while searching for `cmd`
:param bool check: Raise a CommandFailedException if the exit code
is not present in `allowed_retval`
:param list[int] allowed_retval: List of exit codes considered as a
successful termination.
"""
options = []
if ssh:
options += ['-e', self._cmd_quote(ssh, ssh_options)]
if network_compression:
options += ['-z']
# Include patterns must be before the exclude ones, because the exclude
# patterns actually short-circuit the directory traversal stage
# when rsync finds the files to send.
if include:
for pattern in include:
options += ["--include=%s" % (pattern,)]
if exclude:
for pattern in exclude:
options += ["--exclude=%s" % (pattern,)]
if exclude_and_protect:
for pattern in exclude_and_protect:
options += ["--exclude=%s" % (pattern,),
"--filter=P_%s" % (pattern,)]
if args:
options += self._args_for_suse(args)
if bwlimit is not None and bwlimit > 0:
options += ["--bwlimit=%s" % bwlimit]
# By default check is on and the allowed exit code are 0 and 24
if 'check' not in kwargs:
kwargs['check'] = True
if 'allowed_retval' not in kwargs:
kwargs['allowed_retval'] = (0, 24)
Command.__init__(self, rsync, args=options, path=path, **kwargs)
def _args_for_suse(self, args):
"""
Mangle args for SUSE compatibility
See https://bugzilla.opensuse.org/show_bug.cgi?id=898513
"""
# Prepend any argument starting with ':' with a space
# Workaround for SUSE rsync issue
return [' ' + a if a.startswith(':') else a for a in args]
def get_output(self, *args, **kwargs):
"""
Run the command and return the output and the error (if present)
"""
# Prepares args for SUSE
args = self._args_for_suse(args)
# Invoke the base class method
return super(Rsync, self).get_output(*args, **kwargs)
def from_file_list(self, filelist, src, dst, *args, **kwargs):
"""
This method copies filelist from src to dst.
Returns the return code of the rsync command
"""
if 'stdin' in kwargs:
raise TypeError("from_file_list() doesn't support 'stdin' "
"keyword argument")
input_string = ('\n'.join(filelist)).encode('UTF-8')
_logger.debug("from_file_list: %r", filelist)
kwargs['stdin'] = input_string
self.get_output('--files-from=-', src, dst, *args, **kwargs)
return self.ret
class RsyncPgData(Rsync):
"""
This class is a wrapper for rsync, specialised in sync-ing the
Postgres data directory
"""
def __init__(self, rsync='rsync', args=None, **kwargs):
"""
Constructor
:param str rsync: command to run
"""
options = ['-rLKpts', '--delete-excluded', '--inplace']
if args:
options += args
Rsync.__init__(self, rsync, args=options, **kwargs)
class PostgreSQLClient(Command):
"""
Superclass of all the PostgreSQL client commands.
"""
COMMAND = None
def __init__(self,
connection,
command,
version=None,
app_name=None,
path=None,
**kwargs):
"""
Constructor
:param PostgreSQL connection: an object representing
a database connection
:param str command: the command to use
:param Version version: the command version
:param str app_name: the application name to use for the connection
:param str path: additional path for executable retrieval
"""
Command.__init__(self, command, path=path, **kwargs)
if version and version >= Version("9.3"):
# If version of the client is >= 9.3 we use the connection
# string because allows the user to use all the parameters
# supported by the libpq library to create a connection
conn_string = connection.get_connection_string(app_name)
self.args.append("--dbname=%s" % conn_string)
else:
# 9.2 version doesn't support
# connection strings so the 'split' version of the conninfo
# option is used instead.
conn_params = connection.conn_parameters
self.args.append("--host=%s" % conn_params.get('host', None))
self.args.append("--port=%s" % conn_params.get('port', None))
self.args.append("--username=%s" % conn_params.get('user', None))
self.enable_signal_forwarding(signal.SIGINT)
self.enable_signal_forwarding(signal.SIGTERM)
@classmethod
def get_version_info(cls, path=None):
"""
Return a dictionary containing all the info about
the version of the PostgreSQL client
:param str path: the PATH env
"""
if cls.COMMAND is None:
raise NotImplementedError(
"get_version_info cannot be invoked on %s" % cls.__name__)
version_info = dict.fromkeys(('full_path',
'full_version',
'major_version'),
None)
# Get the version string
try:
command = Command(cls.COMMAND, path=path, check=True)
version_info['full_path'] = command.cmd
command("--version")
except CommandFailedException as e:
_logger.debug("Error invoking %s: %s", cls.COMMAND, e)
return version_info
# Parse the full text version
full_version = command.out.strip().split()[-1]
version_info['full_version'] = Version(full_version)
# Extract the major version
version_info['major_version'] = Version(barman.utils.simplify_version(
full_version))
return version_info
class PgBaseBackup(PostgreSQLClient):
"""
Wrapper class for the pg_basebackup system command
"""
COMMAND = 'pg_basebackup'
def __init__(self,
connection,
destination,
command=COMMAND,
version=None,
app_name=None,
bwlimit=None,
tbs_mapping=None,
immediate=False,
check=True,
args=None,
**kwargs):
"""
Constructor
:param PostgreSQL connection: an object representing
a database connection
:param str destination: destination directory path
:param str command: the command to use
:param Version version: the command version
:param str app_name: the application name to use for the connection
:param str bwlimit: bandwidth limit for pg_basebackup
:param Dict[str, str] tbs_mapping: used for tablespace
:param bool immediate: fast checkpoint identifier for pg_basebackup
:param bool check: check if the return value is in the list of
allowed values of the Command obj
:param List[str] args: additional arguments
"""
PostgreSQLClient.__init__(
self,
connection=connection, command=command,
version=version, app_name=app_name,
check=check, **kwargs)
# Set the backup destination
self.args += ['-v', '--no-password', '--pgdata=%s' % destination]
# The tablespace mapping option is repeated once for each tablespace
if tbs_mapping:
for (tbs_source, tbs_destination) in tbs_mapping.items():
self.args.append('--tablespace-mapping=%s=%s' %
(tbs_source, tbs_destination))
# Only global bandwidth limit is supported
if bwlimit is not None and bwlimit > 0:
self.args.append("--max-rate=%s" % bwlimit)
# Immediate checkpoint
if immediate:
self.args.append("--checkpoint=fast")
# Manage additional args
if args:
self.args += args
class PgReceiveXlog(PostgreSQLClient):
"""
Wrapper class for pg_receivexlog
"""
COMMAND = "pg_receivexlog"
def __init__(self,
connection,
destination,
command=COMMAND,
version=None,
app_name=None,
synchronous=False,
check=True,
slot_name=None,
args=None,
**kwargs):
"""
Constructor
:param PostgreSQL connection: an object representing
a database connection
:param str destination: destination directory path
:param str command: the command to use
:param Version version: the command version
:param str app_name: the application name to use for the connection
:param bool synchronous: request synchronous WAL streaming
:param bool check: check if the return value is in the list of
allowed values of the Command obj
:param str slot_name: the replication slot name to use for the
connection
:param List[str] args: additional arguments
"""
PostgreSQLClient.__init__(
self,
connection=connection, command=command,
version=version, app_name=app_name,
check=check, **kwargs)
self.args += [
"--verbose",
"--no-loop",
"--no-password",
"--directory=%s" % destination]
# Add the replication slot name if set in the configuration.
if slot_name is not None:
self.args.append('--slot=%s' % slot_name)
# Request synchronous mode
if synchronous:
self.args.append('--synchronous')
# Manage additional args
if args:
self.args += args
class BarmanSubProcess(object):
"""
Wrapper class for barman sub instances
"""
def __init__(self, command=sys.argv[0], subcommand=None,
config=None, args=None):
"""
Build a specific wrapper for all the barman sub-commands,
providing an unified interface.
:param str command: path to barman
:param str subcommand: the barman sub-command
:param str config: path to the barman configuration file.
:param list[str] args: a list containing the sub-command args
like the target server name
"""
# The config argument is needed when the user explicitly
# passes a configuration file, as the child process
# must know the configuration file to use.
#
# The configuration file must always be propagated,
# even in case of the default one.
if not config:
raise CommandFailedException(
"No configuration file passed to barman subprocess")
# Build the sub-command:
# * be sure to run it with the right python interpreter
# * pass the current configuration file with -c
# * set it quiet with -q
self.command = [sys.executable, command,
'-c', config, '-q', subcommand]
# Handle args for the sub-command (like the server name)
if args:
self.command += args
def execute(self):
"""
Execute the command and pass the output to the configured handlers
"""
_logger.debug("BarmanSubProcess: %r", self.command)
# Redirect all descriptors to /dev/null
devnull = open(os.devnull, 'a+')
proc = subprocess.Popen(
self.command,
preexec_fn=os.setsid, close_fds=True,
stdin=devnull, stdout=devnull, stderr=devnull)
_logger.debug("BarmanSubProcess: subprocess started. pid: %s",
proc.pid)
barman-2.1/barman/recovery_executor.py 0000664 0001750 0001750 00000127403 13016764724 020621 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module contains the methods necessary to perform a recovery
"""
from __future__ import print_function
import collections
import logging
import os
import re
import shutil
import socket
import tempfile
import time
from io import StringIO
import dateutil.parser
import dateutil.tz
from barman import output, xlog
from barman.command_wrappers import RsyncPgData
from barman.config import RecoveryOptions
from barman.copy_controller import RsyncCopyController
from barman.exceptions import (BadXlogSegmentName, CommandFailedException,
DataTransferFailure, FsOperationFailed)
from barman.fs import UnixLocalCommand, UnixRemoteCommand
from barman.infofile import BackupInfo
from barman.utils import mkpath
# generic logger for this module
_logger = logging.getLogger(__name__)
# regexp matching a single value in Postgres configuration file
PG_CONF_SETTING_RE = re.compile(r"^\s*([^\s=]+)\s*=?\s*(.*)$")
# create a namedtuple object called Assertion
# with 'filename', 'line', 'key' and 'value' as properties
Assertion = collections.namedtuple('Assertion', 'filename line key value')
# noinspection PyMethodMayBeStatic
class RecoveryExecutor(object):
"""
Class responsible of recovery operations
"""
# Potentially dangerous options list, which need to be revised by the user
# after a recovery
DANGEROUS_OPTIONS = ['data_directory', 'config_file', 'hba_file',
'ident_file', 'external_pid_file', 'ssl_cert_file',
'ssl_key_file', 'ssl_ca_file', 'ssl_crl_file',
'unix_socket_directory', 'include', 'include_dir',
'include_if_exists']
# List of options that, if present, need to be forced to a specific value
# during recovery, to avoid data losses
MANGLE_OPTIONS = {'archive_command': 'false'}
def __init__(self, backup_manager):
"""
Constructor
:param barman.backup.BackupManager backup_manager: the BackupManager
owner of the executor
"""
self.backup_manager = backup_manager
self.server = backup_manager.server
self.config = backup_manager.config
def recover(self, backup_info, dest, tablespaces, target_tli,
target_time, target_xid, target_name,
exclusive, remote_command):
"""
Performs a recovery of a backup
:param barman.infofile.BackupInfo backup_info: the backup to recover
:param str dest: the destination directory
:param dict[str,str]|None tablespaces: a tablespace
name -> location map (for relocation)
:param str|None target_tli: the target timeline
:param str|None target_time: the target time
:param str|None target_xid: the target xid
:param str|None target_name: the target name created previously with
pg_create_restore_point() function call
:param bool exclusive: whether the recovery is exclusive or not
:param str|None remote_command: The remote command to recover
the base backup, in case of remote backup.
"""
# Run the cron to be sure the wal catalog is up to date
# Prepare a map that contains all the objects required for a recovery
recovery_info = self._setup(backup_info, remote_command, dest)
output.info("Starting %s restore for server %s using backup %s",
recovery_info['recovery_dest'], self.server.config.name,
backup_info.backup_id)
output.info("Destination directory: %s", dest)
# Set targets for PITR
self._set_pitr_targets(recovery_info,
backup_info, dest,
target_name,
target_time,
target_tli,
target_xid)
# Retrieve the safe_horizon for smart copy
self._retrieve_safe_horizon(recovery_info, backup_info, dest)
# check destination directory. If doesn't exist create it
try:
recovery_info['cmd'].create_dir_if_not_exists(dest)
except FsOperationFailed as e:
output.error("unable to initialise destination directory "
"'%s': %s", dest, e)
output.close_and_exit()
# Initialize tablespace directories
if backup_info.tablespaces:
self._prepare_tablespaces(backup_info,
recovery_info['cmd'],
dest,
tablespaces)
# Copy the base backup
output.info("Copying the base backup.")
try:
self._backup_copy(
backup_info, dest,
tablespaces, remote_command,
recovery_info['safe_horizon'])
except DataTransferFailure as e:
output.error("Failure copying base backup: %s", e)
output.close_and_exit()
# Copy the backup.info file in the destination as
# ".barman-recover.info"
if remote_command:
try:
recovery_info['rsync'](backup_info.filename,
':%s/.barman-recover.info' % dest)
except CommandFailedException as e:
output.error(
'copy of recovery metadata file failed: %s', e)
output.close_and_exit()
else:
backup_info.save(os.path.join(dest, '.barman-recover.info'))
# Restore the WAL segments. If GET_WAL option is set, skip this phase
# as they will be retrieved using the wal-get command.
if not recovery_info['get_wal']:
output.info("Copying required WAL segments.")
try:
# Retrieve a list of required log files
required_xlog_files = tuple(
self.server.get_required_xlog_files(
backup_info, target_tli,
recovery_info['target_epoch']))
# Restore WAL segments into the wal_dest directory
self._xlog_copy(required_xlog_files,
recovery_info['wal_dest'],
remote_command)
except DataTransferFailure as e:
output.error("Failure copying WAL files: %s", e)
output.close_and_exit()
except BadXlogSegmentName as e:
output.error(
"invalid xlog segment name %r\n"
"HINT: Please run \"barman rebuild-xlogdb %s\" "
"to solve this issue" %
str(e), self.config.name)
output.close_and_exit()
# If WAL files are put directly in the pg_xlog directory,
# avoid shipping of just recovered files
# by creating the corresponding archive status file
if not recovery_info['is_pitr']:
output.info("Generating archive status files")
self._generate_archive_status(recovery_info,
remote_command,
required_xlog_files)
# Generate recovery.conf file (only if needed by PITR)
if recovery_info['is_pitr']:
output.info("Generating recovery.conf")
self._generate_recovery_conf(recovery_info, backup_info, dest,
exclusive, remote_command,
target_name, target_time,
target_tli, target_xid)
# Create archive_status directory if necessary
archive_status_dir = os.path.join(dest, 'pg_xlog', 'archive_status')
try:
recovery_info['cmd'].create_dir_if_not_exists(archive_status_dir)
except FsOperationFailed as e:
output.error("unable to create the archive_status directory "
"'%s': %s", archive_status_dir, e)
output.close_and_exit()
# As last step, analyse configuration files in order to spot
# harmful options. Barman performs automatic conversion of
# some options as well as notifying users of their existence.
#
# This operation is performed in three steps:
# 1) mapping
# 2) analysis
# 3) copy
output.info("Identify dangerous settings in destination directory.")
self._map_temporary_config_files(recovery_info,
backup_info,
remote_command)
self._analyse_temporary_config_files(recovery_info)
self._copy_temporary_config_files(dest,
remote_command,
recovery_info)
# Cleanup operations
self._teardown(recovery_info)
return recovery_info
def _setup(self, backup_info, remote_command, dest):
"""
Prepare the recovery_info dictionary for the recovery, as well
as temporary working directory
:param barman.infofile.BackupInfo backup_info: representation of a
backup
:param str remote_command: ssh command for remote connection
:return dict: recovery_info dictionary, holding the basic values for a
recovery
"""
recovery_info = {
'cmd': None,
'recovery_dest': 'local',
'rsync': None,
'configuration_files': [],
'destination_path': dest,
'temporary_configuration_files': [],
'tempdir': tempfile.mkdtemp(prefix='barman_recovery-'),
'is_pitr': False,
'wal_dest': os.path.join(dest, 'pg_xlog'),
'get_wal': RecoveryOptions.GET_WAL in self.config.recovery_options,
}
# A map that will keep track of the results of the recovery.
# Used for output generation
results = {
'changes': [],
'warnings': [],
'delete_barman_xlog': False,
'missing_files': [],
'get_wal': False,
}
recovery_info['results'] = results
# Set up a list of configuration files
recovery_info['configuration_files'].append('postgresql.conf')
if backup_info.version >= 90400:
recovery_info['configuration_files'].append('postgresql.auto.conf')
# Handle remote recovery options
if remote_command:
recovery_info['recovery_dest'] = 'remote'
try:
recovery_info['rsync'] = RsyncPgData(
path=self.server.path,
ssh=remote_command,
bwlimit=self.config.bandwidth_limit,
network_compression=self.config.network_compression)
except CommandFailedException:
self._teardown(recovery_info)
raise
try:
# create a UnixRemoteCommand obj if is a remote recovery
recovery_info['cmd'] = UnixRemoteCommand(remote_command,
path=self.server.path)
except FsOperationFailed:
self._teardown(recovery_info)
output.error(
"Unable to connect to the target host using the command "
"'%s'", remote_command)
output.close_and_exit()
else:
# if is a local recovery create a UnixLocalCommand
recovery_info['cmd'] = UnixLocalCommand()
return recovery_info
def _set_pitr_targets(self, recovery_info, backup_info, dest, target_name,
target_time, target_tli, target_xid):
"""
Set PITR targets - as specified by the user
:param dict recovery_info: Dictionary containing all the recovery
parameters
:param barman.infofile.BackupInfo backup_info: representation of a
backup
:param str dest: destination directory of the recovery
:param str|None target_name: recovery target name for PITR
:param str|None target_time: recovery target time for PITR
:param str|None target_tli: recovery target timeline for PITR
:param str|None target_xid: recovery target transaction id for PITR
"""
target_epoch = None
target_datetime = None
if (target_time or
target_xid or
(target_tli and target_tli != backup_info.timeline) or
target_name or
recovery_info['get_wal']):
recovery_info['is_pitr'] = True
targets = {}
if target_time:
# noinspection PyBroadException
try:
target_datetime = dateutil.parser.parse(target_time)
except ValueError as e:
output.error(
"unable to parse the target time parameter %r: %s",
target_time, e)
self._teardown(recovery_info)
output.close_and_exit()
except Exception:
# this should not happen, but there is a known bug in
# dateutil.parser.parse() implementation
# ref: https://bugs.launchpad.net/dateutil/+bug/1247643
output.error(
"unable to parse the target time parameter %r",
target_time)
output.close_and_exit()
target_epoch = (
time.mktime(target_datetime.timetuple()) +
(target_datetime.microsecond / 1000000.))
targets['time'] = str(target_datetime)
if target_xid:
targets['xid'] = str(target_xid)
if target_tli and target_tli != backup_info.timeline:
targets['timeline'] = str(target_tli)
if target_name:
targets['name'] = str(target_name)
output.info(
"Doing PITR. Recovery target %s",
(", ".join(["%s: %r" % (k, v) for k, v in targets.items()])))
recovery_info['wal_dest'] = os.path.join(dest, 'barman_xlog')
# With a PostgreSQL version older than 8.4, it is the user's
# responsibility to delete the "barman_xlog" directory as the
# restore_command option in recovery.conf is not supported
if backup_info.version < 80400 and \
not recovery_info['get_wal']:
recovery_info['results']['delete_barman_xlog'] = True
recovery_info['target_epoch'] = target_epoch
recovery_info['target_datetime'] = target_datetime
def _retrieve_safe_horizon(self, recovery_info, backup_info, dest):
"""
Retrieve the safe_horizon for smart copy
If the target directory contains a previous recovery, it is safe to
pick the least of the two backup "begin times" (the one we are
recovering now and the one previously recovered in the target
directory). Set the value in the given recovery_info dictionary.
:param dict recovery_info: Dictionary containing all the recovery
parameters
:param barman.infofile.BackupInfo backup_info: a backup representation
:param str dest: recovery destination directory
"""
# noinspection PyBroadException
try:
backup_begin_time = backup_info.begin_time
# Retrieve previously recovered backup metadata (if available)
dest_info_txt = recovery_info['cmd'].get_file_content(
os.path.join(dest, '.barman-recover.info'))
dest_info = BackupInfo(
self.server,
info_file=StringIO(dest_info_txt))
dest_begin_time = dest_info.begin_time
# Pick the earlier begin time. Both are tz-aware timestamps because
# BackupInfo class ensure it
safe_horizon = min(backup_begin_time, dest_begin_time)
output.info("Using safe horizon time for smart rsync copy: %s",
safe_horizon)
except FsOperationFailed as e:
# Setting safe_horizon to None will effectively disable
# the time-based part of smart_copy method. However it is still
# faster than running all the transfers with checksum enabled.
#
# FsOperationFailed means the .barman-recover.info is not available
# on destination directory
safe_horizon = None
_logger.warning('Unable to retrieve safe horizon time '
'for smart rsync copy: %s', e)
except Exception as e:
# Same as above, but something failed decoding .barman-recover.info
# or comparing times, so log the full traceback
safe_horizon = None
_logger.exception('Error retrieving safe horizon time '
'for smart rsync copy: %s', e)
recovery_info['safe_horizon'] = safe_horizon
def _prepare_tablespaces(self, backup_info, cmd, dest, tablespaces):
"""
Prepare the directory structure for required tablespaces,
taking care of tablespaces relocation, if requested.
:param barman.infofile.BackupInfo backup_info: backup representation
:param barman.fs.UnixLocalCommand cmd: Object for
filesystem interaction
:param str dest: destination dir for the recovery
:param dict tablespaces: dict of all the tablespaces and their location
"""
tblspc_dir = os.path.join(dest, 'pg_tblspc')
try:
# check for pg_tblspc dir into recovery destination folder.
# if it does not exists, create it
cmd.create_dir_if_not_exists(tblspc_dir)
except FsOperationFailed as e:
output.error("unable to initialise tablespace directory "
"'%s': %s", tblspc_dir, e)
output.close_and_exit()
for item in backup_info.tablespaces:
# build the filename of the link under pg_tblspc directory
pg_tblspc_file = os.path.join(tblspc_dir, str(item.oid))
# by default a tablespace goes in the same location where
# it was on the source server when the backup was taken
location = item.location
# if a relocation has been requested for this tablespace,
# use the target directory provided by the user
if tablespaces and item.name in tablespaces:
location = tablespaces[item.name]
try:
# remove the current link in pg_tblspc, if it exists
cmd.delete_if_exists(pg_tblspc_file)
# create tablespace location, if does not exist
# (raise an exception if it is not possible)
cmd.create_dir_if_not_exists(location)
# check for write permissions on destination directory
cmd.check_write_permission(location)
# create symlink between tablespace and recovery folder
cmd.create_symbolic_link(location, pg_tblspc_file)
except FsOperationFailed as e:
output.error("unable to prepare '%s' tablespace "
"(destination '%s'): %s",
item.name, location, e)
output.close_and_exit()
output.info("\t%s, %s, %s", item.oid, item.name, location)
def _backup_copy(self, backup_info, dest, tablespaces=None,
remote_command=None, safe_horizon=None):
"""
Perform the actual copy of the base backup for recovery purposes
First, it copies one tablespace at a time, then the PGDATA directory.
Bandwidth limitation, according to configuration, is applied in
the process.
TODO: manage configuration files if outside PGDATA.
:param barman.infofile.BackupInfo backup_info: the backup to recover
:param str dest: the destination directory
:param dict[str,str]|None tablespaces: a tablespace
name -> location map (for relocation)
:param str|None remote_command: default None. The remote command to
recover the base backup, in case of remote backup.
:param datetime.datetime|None safe_horizon: anything after this time
has to be checked with checksum
"""
# Set a ':' prefix to remote destinations
dest_prefix = ''
if remote_command:
dest_prefix = ':'
# Create the copy controller object, specific for rsync,
# which will drive all the copy operations. Items to be
# copied are added before executing the copy() method
controller = RsyncCopyController(
path=self.server.path,
ssh_command=remote_command,
network_compression=self.config.network_compression,
safe_horizon=safe_horizon,
retry_times=self.config.basebackup_retry_times,
retry_sleep=self.config.basebackup_retry_sleep,
)
# Dictionary for paths to be excluded from rsync
exclude_and_protect = []
# Process every tablespace
if backup_info.tablespaces:
for tablespace in backup_info.tablespaces:
# By default a tablespace goes in the same location where
# it was on the source server when the backup was taken
location = tablespace.location
# If a relocation has been requested for this tablespace
# use the user provided target directory
if tablespaces and tablespace.name in tablespaces:
location = tablespaces[tablespace.name]
# If the tablespace location is inside the data directory,
# exclude and protect it from being deleted during
# the data directory copy
if location.startswith(dest):
exclude_and_protect += [location[len(dest):]]
# Exclude and protect the tablespace from being deleted during
# the data directory copy
exclude_and_protect.append("/pg_tblspc/%s" % tablespace.oid)
# Add the tablespace directory to the list of objects
# to be copied by the controller
controller.add_directory(
label=tablespace.name,
src='%s/' % backup_info.get_data_directory(tablespace.oid),
dst=dest_prefix + location,
bwlimit=self.config.get_bwlimit(tablespace),
item_class=controller.TABLESPACE_CLASS
)
# Add the PGDATA directory to the list of objects to be copied
# by the controller
controller.add_directory(
label='pgdata',
src='%s/' % backup_info.get_data_directory(),
dst=dest_prefix + dest,
bwlimit=self.config.get_bwlimit(),
exclude=[
'/pg_xlog/*',
'/pg_log/*',
'/recovery.conf',
'/postmaster.pid'],
exclude_and_protect=exclude_and_protect,
item_class=controller.PGDATA_CLASS
)
# TODO: Manage different location for configuration files
# TODO: that were not within the data directory
# Execute the copy
try:
controller.copy()
# TODO: Improve the exception output
except CommandFailedException as e:
msg = "data transfer failure"
raise DataTransferFailure.from_command_error(
'rsync', e, msg)
def _xlog_copy(self, required_xlog_files, wal_dest, remote_command):
"""
Restore WAL segments
:param required_xlog_files: list of all required WAL files
:param wal_dest: the destination directory for xlog recover
:param remote_command: default None. The remote command to recover
the xlog, in case of remote backup.
"""
# List of required WAL files partitioned by containing directory
xlogs = collections.defaultdict(list)
# add '/' suffix to ensure it is a directory
wal_dest = '%s/' % wal_dest
# Map of every compressor used with any WAL file in the archive,
# to be used during this recovery
compressors = {}
compression_manager = self.backup_manager.compression_manager
# Fill xlogs and compressors maps from required_xlog_files
for wal_info in required_xlog_files:
hashdir = xlog.hash_dir(wal_info.name)
xlogs[hashdir].append(wal_info)
# If a compressor is required, make sure it exists in the cache
if wal_info.compression is not None and \
wal_info.compression not in compressors:
compressors[wal_info.compression] = \
compression_manager.get_compressor(
compression=wal_info.compression)
rsync = RsyncPgData(
path=self.server.path,
ssh=remote_command,
bwlimit=self.config.bandwidth_limit,
network_compression=self.config.network_compression)
# If compression is used and this is a remote recovery, we need a
# temporary directory where to spool uncompressed files,
# otherwise we either decompress every WAL file in the local
# destination, or we ship the uncompressed file remotely
if compressors:
if remote_command:
# Decompress to a temporary spool directory
wal_decompression_dest = tempfile.mkdtemp(
prefix='barman_xlog-')
else:
# Decompress directly to the destination directory
wal_decompression_dest = wal_dest
# Make sure wal_decompression_dest exists
mkpath(wal_decompression_dest)
else:
# If no compression
wal_decompression_dest = None
if remote_command:
# If remote recovery tell rsync to copy them remotely
# add ':' prefix to mark it as remote
wal_dest = ':%s' % wal_dest
total_wals = sum(map(len, xlogs.values()))
partial_count = 0
for prefix in sorted(xlogs):
batch_len = len(xlogs[prefix])
partial_count += batch_len
source_dir = os.path.join(self.config.wals_directory, prefix)
_logger.info(
"Starting copy of %s WAL files %s/%s from %s to %s",
batch_len,
partial_count,
total_wals,
xlogs[prefix][0],
xlogs[prefix][-1])
# If at least one compressed file has been found, activate
# compression check and decompression for each WAL files
if compressors:
for segment in xlogs[prefix]:
dst_file = os.path.join(wal_decompression_dest,
segment.name)
if segment.compression is not None:
compressors[segment.compression].decompress(
os.path.join(source_dir, segment.name),
dst_file)
else:
shutil.copy2(os.path.join(source_dir, segment.name),
dst_file)
if remote_command:
try:
# Transfer the WAL files
rsync.from_file_list(
list(segment.name for segment in xlogs[prefix]),
wal_decompression_dest, wal_dest)
except CommandFailedException as e:
msg = ("data transfer failure while copying WAL files "
"to directory '%s'") % (wal_dest[1:],)
raise DataTransferFailure.from_command_error(
'rsync', e, msg)
# Cleanup files after the transfer
for segment in xlogs[prefix]:
file_name = os.path.join(wal_decompression_dest,
segment.name)
try:
os.unlink(file_name)
except OSError as e:
output.warning(
"Error removing temporary file '%s': %s",
file_name, e)
else:
try:
rsync.from_file_list(
list(segment.name for segment in xlogs[prefix]),
"%s/" % os.path.join(self.config.wals_directory,
prefix),
wal_dest)
except CommandFailedException as e:
msg = "data transfer failure while copying WAL files " \
"to directory '%s'" % (wal_dest[1:],)
raise DataTransferFailure.from_command_error(
'rsync', e, msg)
_logger.info("Finished copying %s WAL files.", total_wals)
# Remove local decompression target directory if different from the
# destination directory (it happens when compression is in use during a
# remote recovery
if wal_decompression_dest and wal_decompression_dest != wal_dest:
shutil.rmtree(wal_decompression_dest)
def _generate_archive_status(self, recovery_info, remote_command,
required_xlog_files):
"""
Populate the archive_status directory
:param dict recovery_info: Dictionary containing all the recovery
parameters
:param str remote_command: ssh command for remote connection
:param tuple required_xlog_files: list of required WAL segments
"""
if remote_command:
status_dir = recovery_info['tempdir']
else:
status_dir = os.path.join(recovery_info['wal_dest'],
'archive_status')
mkpath(status_dir)
for wal_info in required_xlog_files:
with open(os.path.join(status_dir, "%s.done" % wal_info.name),
'a') as f:
f.write('')
if remote_command:
try:
recovery_info['rsync']('%s/' % status_dir,
':%s' % os.path.join(
recovery_info['wal_dest'],
'archive_status'))
except CommandFailedException as e:
output.error("unable to populate pg_xlog/archive_status "
"directory: %s", e)
output.close_and_exit()
def _generate_recovery_conf(self, recovery_info, backup_info, dest,
exclusive, remote_command, target_name,
target_time, target_tli, target_xid):
"""
Generate a recovery.conf file for PITR containing
all the required configurations
:param dict recovery_info: Dictionary containing all the recovery
parameters
:param barman.infofile.BackupInfo backup_info: representation of a
backup
:param str dest: destination directory of the recovery
:param boolean exclusive: exclusive backup or concurrent
:param str remote_command: ssh command for remote connection
:param str target_name: recovery target name for PITR
:param str target_time: recovery target time for PITR
:param str target_tli: recovery target timeline for PITR
:param str target_xid: recovery target transaction id for PITR
"""
if remote_command:
recovery = open(os.path.join(recovery_info['tempdir'],
'recovery.conf'), 'w')
else:
recovery = open(os.path.join(dest, 'recovery.conf'), 'w')
# If GET_WAL has been set, use the get-wal command to retrieve the
# required wal files. Otherwise use the unix command "cp" to copy
# them from the barman_xlog directory
if recovery_info['get_wal']:
# We need to create the right restore command.
# If we are doing a remote recovery,
# the barman-cli package is REQUIRED on the server that is hosting
# the PostgreSQL server.
# We use the machine FQDN and the barman_user
# setting to call the barman-wal-restore correctly.
# If local recovery, we use barman directly, assuming
# the postgres process will be executed with the barman user.
# It MUST to be reviewed by the user in any case.
if remote_command:
fqdn = socket.getfqdn()
print("# The 'barman-wal-restore' command "
"is provided in the 'barman-cli' package",
file=recovery)
print("restore_command = 'barman-wal-restore -U %s "
"%s %s %%f %%p'" % (self.config.config.user,
fqdn, self.config.name),
file=recovery)
else:
print("# The 'barman get-wal' command "
"must run as '%s' user" % self.config.config.user,
file=recovery)
print("restore_command = 'sudo -u %s "
"barman get-wal %s %%f > %%p'" % (
self.config.config.user,
self.config.name),
file=recovery)
recovery_info['results']['get_wal'] = True
else:
print("restore_command = 'cp barman_xlog/%f %p'", file=recovery)
if backup_info.version >= 80400 and \
not recovery_info['get_wal']:
print("recovery_end_command = 'rm -fr barman_xlog'", file=recovery)
if target_time:
print("recovery_target_time = '%s'" % target_time, file=recovery)
if target_tli:
print("recovery_target_timeline = %s" % target_tli, file=recovery)
if target_xid:
print("recovery_target_xid = '%s'" % target_xid, file=recovery)
if target_name:
print("recovery_target_name = '%s'" % target_name, file=recovery)
if (target_xid or target_time) and exclusive:
print("recovery_target_inclusive = '%s'" % (
not exclusive), file=recovery)
recovery.close()
if remote_command:
plain_rsync = RsyncPgData(
path=self.server.path,
ssh=remote_command,
bwlimit=self.config.bandwidth_limit,
network_compression=self.config.network_compression)
try:
plain_rsync.from_file_list(['recovery.conf'],
recovery_info['tempdir'],
':%s' % dest)
except CommandFailedException as e:
output.error('remote copy of recovery.conf failed: %s', e)
output.close_and_exit()
def _map_temporary_config_files(self, recovery_info, backup_info,
remote_command):
"""
Map configuration files, by filling the 'temporary_configuration_files'
array, depending on remote or local recovery. This array will be used
by the subsequent methods of the class.
:param dict recovery_info: Dictionary containing all the recovery
parameters
:param barman.infofile.BackupInfo backup_info: a backup representation
:param str remote_command: ssh command for remote recovery
"""
# Cycle over postgres configuration files which my be missing.
# If a file is missing, we will be unable to restore it and
# we will warn the user.
# This can happen if we are using pg_basebackup and
# a configuration file is located outside the data dir.
# This is not an error condition, so we check also for
# `pg_ident.conf` which is an optional file.
for conf_file in (recovery_info['configuration_files'] +
['pg_hba.conf', 'pg_ident.conf']):
source_path = os.path.join(
backup_info.get_data_directory(), conf_file)
if not os.path.exists(source_path):
recovery_info['results']['missing_files'].append(conf_file)
# Remove the file from the list of configuration files
if conf_file in recovery_info['configuration_files']:
recovery_info['configuration_files'].remove(conf_file)
for conf_file in recovery_info['configuration_files']:
if remote_command:
# If the recovery is remote, copy the postgresql.conf
# file in a temp dir
# Otherwise we can modify the postgresql.conf file
# in the destination directory.
conf_file_path = os.path.join(
recovery_info['tempdir'], conf_file)
shutil.copy2(
os.path.join(backup_info.get_data_directory(),
conf_file), conf_file_path)
else:
# Otherwise use the local destination path.
conf_file_path = os.path.join(
recovery_info['destination_path'], conf_file)
recovery_info['temporary_configuration_files'].append(
conf_file_path)
def _analyse_temporary_config_files(self, recovery_info):
"""
Analyse temporary configuration files and identify dangerous options
Mark all the dangerous options for the user to review. This procedure
also changes harmful options such as 'archive_command'.
:param dict recovery_info: dictionary holding all recovery parameters
"""
results = recovery_info['results']
# Check for dangerous options inside every config file
for conf_file in recovery_info['temporary_configuration_files']:
# Identify and comment out dangerous options, replacing them with
# the appropriate values
results['changes'] += self._pg_config_mangle(
conf_file,
self.MANGLE_OPTIONS,
"%s.origin" % conf_file)
# Identify dangerous options and warn users about their presence
results['warnings'] += self._pg_config_detect_possible_issues(
conf_file)
def _copy_temporary_config_files(self, dest,
remote_command, recovery_info):
"""
Copy modified configuration files using rsync in case of
remote recovery
:param str dest: destination directory of the recovery
:param str remote_command: ssh command for remote connection
:param dict recovery_info: Dictionary containing all the recovery
parameters
"""
if remote_command:
# If this is a remote recovery, rsync the modified files from the
# temporary local directory to the remote destination directory.
file_list = []
for conf_file in recovery_info['configuration_files']:
file_list.append('%s' % conf_file)
file_list.append('%s.origin' % conf_file)
try:
recovery_info['rsync'].from_file_list(file_list,
recovery_info['tempdir'],
':%s' % dest)
except CommandFailedException as e:
output.error('remote copy of configuration files failed: %s',
e)
output.close_and_exit()
def _teardown(self, recovery_info):
"""
Cleanup operations for a recovery
:param dict recovery_info: dictionary holding the basic values
for a recovery
"""
# Remove the temporary directory (created in the setup method)
shutil.rmtree(recovery_info['tempdir'])
def _pg_config_mangle(self, filename, settings, backup_filename=None):
"""
This method modifies the given PostgreSQL configuration file,
commenting out the given settings, and adding the ones generated by
Barman.
If backup_filename is passed, keep a backup copy.
:param filename: the PostgreSQL configuration file
:param settings: dictionary of settings to be mangled
:param backup_filename: config file backup copy. Default is None.
"""
# Read the full content of the file in memory
with open(filename) as f:
content = f.readlines()
# Rename the original file to backup_filename or to a temporary name
# if backup_filename is missing. We need to keep it to preserve
# permissions.
if backup_filename:
orig_filename = backup_filename
else:
orig_filename = "%s.config_mangle.old" % filename
shutil.move(filename, orig_filename)
# Write the mangled content
mangled = []
with open(filename, 'w') as f:
for l_number, line in enumerate(content):
rm = PG_CONF_SETTING_RE.match(line)
if rm:
key = rm.group(1)
if key in settings:
f.write("#BARMAN# %s" % line)
# TODO is it useful to handle none values?
changes = "%s = %s\n" % (key, settings[key])
f.write(changes)
mangled.append(
Assertion._make([
os.path.basename(f.name),
l_number,
key,
settings[key]]))
continue
f.write(line)
# Restore original permissions
shutil.copymode(orig_filename, filename)
# If a backup copy of the file is not requested,
# unlink the orig file
if not backup_filename:
os.unlink(orig_filename)
return mangled
def _pg_config_detect_possible_issues(self, filename):
"""
This method looks for any possible issue with PostgreSQL
location options such as data_directory, config_file, etc.
It returns a dictionary with the dangerous options that
have been found.
:param filename: the Postgres configuration file
"""
clashes = []
with open(filename) as f:
content = f.readlines()
# Read line by line and identify dangerous options
for l_number, line in enumerate(content):
rm = PG_CONF_SETTING_RE.match(line)
if rm:
key = rm.group(1)
if key in self.DANGEROUS_OPTIONS:
clashes.append(
Assertion._make([
os.path.basename(f.name),
l_number,
key,
rm.group(2)]))
return clashes
barman-2.1/barman/retention_policies.py 0000664 0001750 0001750 00000034260 12726221047 020732 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module defines backup retention policies. A backup retention
policy in Barman is a user-defined policy for determining how long
backups and archived logs (WAL segments) need to be retained for media
recovery.
You can define a retention policy in terms of backup redundancy
or a recovery window.
Barman retains the periodical backups required to satisfy
the current retention policy, and any archived WAL files required for complete
recovery of those backups.
"""
import logging
import re
from abc import ABCMeta, abstractmethod
from datetime import datetime, timedelta
from dateutil import tz
from barman.infofile import BackupInfo
from barman.utils import with_metaclass
_logger = logging.getLogger(__name__)
class RetentionPolicy(with_metaclass(ABCMeta, object)):
"""Abstract base class for retention policies"""
def __init__(self, mode, unit, value, context, server):
"""Constructor of the retention policy base class"""
self.mode = mode
self.unit = unit
self.value = int(value)
self.context = context
self.server = server
self._first_backup = None
self._first_wal = None
def report(self, source=None, context=None):
"""Report obsolete/valid objects according to the retention policy"""
if context is None:
context = self.context
# Overrides the list of available backups
if source is None:
source = self.server.get_available_backups(
BackupInfo.STATUS_NOT_EMPTY)
if context == 'BASE':
return self._backup_report(source)
elif context == 'WAL':
return self._wal_report()
else:
raise ValueError('Invalid context %s', context)
def backup_status(self, backup_id):
"""Report the status of a backup according to the retention policy"""
source = self.server.get_available_backups(BackupInfo.STATUS_NOT_EMPTY)
if self.context == 'BASE':
return self._backup_report(source)[backup_id]
else:
return BackupInfo.NONE
def first_backup(self):
"""Returns the first valid backup according to retention policies"""
if not self._first_backup:
self.report(context='BASE')
return self._first_backup
def first_wal(self):
"""Returns the first valid WAL according to retention policies"""
if not self._first_wal:
self.report(context='WAL')
return self._first_wal
@abstractmethod
def __str__(self):
"""String representation"""
pass
@abstractmethod
def debug(self):
"""Debug information"""
pass
@abstractmethod
def _backup_report(self, source):
"""Report obsolete/valid backups according to the retention policy"""
pass
@abstractmethod
def _wal_report(self):
"""Report obsolete/valid WALs according to the retention policy"""
pass
@classmethod
def create(cls, server, option, value):
"""
If given option and value from the configuration file match,
creates the retention policy object for the given server
"""
# using @abstractclassmethod from python3 would be better here
raise NotImplementedError(
'The class %s must override the create() class method',
cls.__name__)
def to_json(self):
"""
Output representation of the obj for JSON serialization
"""
return "%s %s %s" % (self.mode, self.value, self.unit)
class RedundancyRetentionPolicy(RetentionPolicy):
"""
Retention policy based on redundancy, the setting that determines
many periodical backups to keep. A redundancy-based retention policy
is contrasted with retention policy that uses a recovery window.
"""
_re = re.compile(r'^\s*redundancy\s+(\d+)\s*$', re.IGNORECASE)
def __init__(self, context, value, server):
super(RedundancyRetentionPolicy, self
).__init__('redundancy', 'b', value, 'BASE', server)
assert (value >= 0)
def __str__(self):
return "REDUNDANCY %s" % self.value
def debug(self):
return "Redundancy: %s (%s)" % (self.value, self.context)
def _backup_report(self, source):
"""Report obsolete/valid backups according to the retention policy"""
report = dict()
backups = source
# Normalise the redundancy value (according to minimum redundancy)
redundancy = self.value
if redundancy < self.server.config.minimum_redundancy:
_logger.warning(
"Retention policy redundancy (%s) is lower than "
"the required minimum redundancy (%s). Enforce %s.",
redundancy, self.server.config.minimum_redundancy,
self.server.config.minimum_redundancy)
redundancy = self.server.config.minimum_redundancy
# Map the latest 'redundancy' DONE backups as VALID
# The remaining DONE backups are classified as OBSOLETE
# Non DONE backups are classified as NONE
# NOTE: reverse key orders (simulate reverse chronology)
i = 0
for bid in sorted(backups.keys(), reverse=True):
if backups[bid].status == BackupInfo.DONE:
if i < redundancy:
report[bid] = BackupInfo.VALID
self._first_backup = bid
else:
report[bid] = BackupInfo.OBSOLETE
i = i + 1
else:
report[bid] = BackupInfo.NONE
return report
def _wal_report(self):
"""Report obsolete/valid WALs according to the retention policy"""
pass
@classmethod
def create(cls, server, context, optval):
# Detect Redundancy retention type
mtch = cls._re.match(optval)
if not mtch:
return None
value = int(mtch.groups()[0])
return cls(context, value, server)
class RecoveryWindowRetentionPolicy(RetentionPolicy):
"""
Retention policy based on recovery window. The DBA specifies a period of
time and Barman ensures retention of backups and archived WAL files
required for point-in-time recovery to any time during the recovery window.
The interval always ends with the current time and extends back in time
for the number of days specified by the user.
For example, if the retention policy is set for a recovery window of
seven days, and the current time is 9:30 AM on Friday, Barman retains
the backups required to allow point-in-time recovery back to 9:30 AM
on the previous Friday.
"""
_re = re.compile(
r"""
^\s*
recovery\s+window\s+of\s+ # recovery window of
(\d+)\s+(day|month|week)s? # N (day|month|week) with optional 's'
\s*$
""",
re.IGNORECASE | re.VERBOSE)
_kw = {'d': 'DAYS', 'm': 'MONTHS', 'w': 'WEEKS'}
def __init__(self, context, value, unit, server):
super(RecoveryWindowRetentionPolicy, self
).__init__('window', unit, value, context, server)
assert (value >= 0)
assert (unit == 'd' or unit == 'm' or unit == 'w')
assert (context == 'WAL' or context == 'BASE')
# Calculates the time delta
if unit == 'd':
self.timedelta = timedelta(days=self.value)
elif unit == 'w':
self.timedelta = timedelta(weeks=self.value)
elif unit == 'm':
self.timedelta = timedelta(days=(31 * self.value))
def __str__(self):
return "RECOVERY WINDOW OF %s %s" % (self.value, self._kw[self.unit])
def debug(self):
return "Recovery Window: %s %s: %s (%s)" % (
self.value, self.unit, self.context,
self._point_of_recoverability())
def _point_of_recoverability(self):
"""
Based on the current time and the window, calculate the point
of recoverability, which will be then used to define the first
backup or the first WAL
"""
return datetime.now(tz.tzlocal()) - self.timedelta
def _backup_report(self, source):
"""Report obsolete/valid backups according to the retention policy"""
report = dict()
backups = source
# Map as VALID all DONE backups having end time lower than
# the point of recoverability. The older ones
# are classified as OBSOLETE.
# Non DONE backups are classified as NONE
found = False
valid = 0
# NOTE: reverse key orders (simulate reverse chronology)
for bid in sorted(backups.keys(), reverse=True):
# We are interested in DONE backups only
if backups[bid].status == BackupInfo.DONE:
if found:
# Check minimum redundancy requirements
if valid < self.server.config.minimum_redundancy:
_logger.warning(
"Keeping obsolete backup %s for server %s "
"(older than %s) "
"due to minimum redundancy requirements (%s)",
bid, self.server.config.name,
self._point_of_recoverability(),
self.server.config.minimum_redundancy)
# We mark the backup as potentially obsolete
# as we must respect minimum redundancy requirements
report[bid] = BackupInfo.POTENTIALLY_OBSOLETE
self._first_backup = bid
valid = valid + 1
else:
# We mark this backup as obsolete
# (older than the first valid one)
_logger.info(
"Reporting backup %s for server %s as OBSOLETE "
"(older than %s)",
bid, self.server.config.name,
self._point_of_recoverability())
report[bid] = BackupInfo.OBSOLETE
else:
_logger.debug(
"Reporting backup %s for server %s as VALID "
"(newer than %s)",
bid, self.server.config.name,
self._point_of_recoverability())
# Backup within the recovery window
report[bid] = BackupInfo.VALID
self._first_backup = bid
valid = valid + 1
# TODO: Currently we use the backup local end time
# We need to make this more accurate
if backups[bid].end_time < self._point_of_recoverability():
found = True
else:
report[bid] = BackupInfo.NONE
return report
def _wal_report(self):
"""Report obsolete/valid WALs according to the retention policy"""
pass
@classmethod
def create(cls, server, context, optval):
# Detect Recovery Window retention type
match = cls._re.match(optval)
if not match:
return None
value = int(match.groups()[0])
unit = match.groups()[1][0].lower()
return cls(context, value, unit, server)
class SimpleWALRetentionPolicy(RetentionPolicy):
"""Simple retention policy for WAL files (identical to the main one)"""
_re = re.compile(r'^\s*main\s*$', re.IGNORECASE)
def __init__(self, context, policy, server):
super(SimpleWALRetentionPolicy, self
).__init__('simple-wal', policy.unit, policy.value,
context, server)
# The referred policy must be of type 'BASE'
assert (self.context == 'WAL' and policy.context == 'BASE')
self.policy = policy
def __str__(self):
return "MAIN"
def debug(self):
return "Simple WAL Retention Policy (%s)" % self.policy
def _backup_report(self, source):
"""Report obsolete/valid backups according to the retention policy"""
pass
def _wal_report(self):
"""Report obsolete/valid backups according to the retention policy"""
self.policy.report(context='WAL')
def first_wal(self):
"""Returns the first valid WAL according to retention policies"""
return self.policy.first_wal()
@classmethod
def create(cls, server, context, optval):
# Detect Redundancy retention type
match = cls._re.match(optval)
if not match:
return None
return cls(context, server.config.retention_policy, server)
class RetentionPolicyFactory(object):
"""Factory for retention policy objects"""
# Available retention policy types
policy_classes = [
RedundancyRetentionPolicy,
RecoveryWindowRetentionPolicy,
SimpleWALRetentionPolicy
]
@classmethod
def create(cls, server, option, value):
"""
Based on the given option and value from the configuration
file, creates the appropriate retention policy object
for the given server
"""
if option == 'wal_retention_policy':
context = 'WAL'
elif option == 'retention_policy':
context = 'BASE'
else:
raise ValueError('Unknown option for retention policy: %s' %
option)
# Look for the matching rule
for policy_class in cls.policy_classes:
policy = policy_class.create(server, context, value)
if policy:
return policy
raise ValueError('Cannot parse option %s: %s' % (option, value))
barman-2.1/barman/diagnose.py 0000664 0001750 0001750 00000006337 13025742160 016626 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module represents the barman diagnostic tool.
"""
import json
import logging
import sys
import barman
from barman import fs, output
from barman.backup import BackupInfo
from barman.exceptions import FsOperationFailed
from barman.utils import BarmanEncoder
_logger = logging.getLogger(__name__)
def exec_diagnose(servers, errors_list):
"""
Diagnostic command: gathers information from backup server
and from all the configured servers.
Gathered information should be used for support and problems detection
:param dict(str,barman.server.Server) servers: list of configured servers
:param list errors_list: list of global errors
"""
# global section. info about barman server
diagnosis = {'global': {}, 'servers': {}}
# barman global config
diagnosis['global']['config'] = dict(barman.__config__._global_config)
diagnosis['global']['config']['errors_list'] = errors_list
command = fs.UnixLocalCommand()
# basic system info
diagnosis['global']['system_info'] = command.get_system_info()
diagnosis['global']['system_info']['barman_ver'] = barman.__version__
# per server section
for name in sorted(servers):
server = servers[name]
if server is None:
output.error("Unknown server '%s'" % name)
continue
# server configuration
diagnosis['servers'][name] = {}
diagnosis['servers'][name]['config'] = vars(server.config)
del diagnosis['servers'][name]['config']['config']
# server system info
if server.config.ssh_command:
try:
command = fs.UnixRemoteCommand(
ssh_command=server.config.ssh_command,
path=server.path
)
diagnosis['servers'][name]['system_info'] = (
command.get_system_info())
except FsOperationFailed:
pass
# barman statuts information for the server
diagnosis['servers'][name]['status'] = server.get_remote_status()
# backup list
backups = server.get_available_backups(BackupInfo.STATUS_ALL)
diagnosis['servers'][name]['backups'] = backups
# wal status
diagnosis['servers'][name]['wals'] = {
'last_archived_wal_per_timeline':
server.backup_manager.get_latest_archived_wals_info(),
}
# Release any PostgreSQL resource
server.close()
output.info(json.dumps(diagnosis, sys.stdout, cls=BarmanEncoder, indent=4,
sort_keys=True))
barman-2.1/barman/server.py 0000664 0001750 0001750 00000233433 13025742160 016342 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module represents a Server.
Barman is able to manage multiple servers.
"""
import itertools
import logging
import os
import shutil
import sys
import time
from collections import namedtuple
from contextlib import contextmanager
from tempfile import NamedTemporaryFile
import barman
from barman import output, xlog
from barman.backup import BackupManager
from barman.command_wrappers import BarmanSubProcess
from barman.compression import identify_compression
from barman.exceptions import (ArchiverFailure, BadXlogSegmentName,
ConninfoException, LockFileBusy,
LockFilePermissionDenied,
PostgresDuplicateReplicationSlot,
PostgresException,
PostgresInvalidReplicationSlot,
PostgresIsInRecovery,
PostgresReplicationSlotInUse,
PostgresReplicationSlotsFull,
PostgresSuperuserRequired,
PostgresUnsupportedFeature, TimeoutError,
UnknownBackupIdException)
from barman.infofile import BackupInfo, WalFileInfo
from barman.lockfile import (ServerBackupLock, ServerCronLock,
ServerWalArchiveLock, ServerWalReceiveLock,
ServerXLOGDBLock)
from barman.postgres import PostgreSQLConnection, StreamingConnection
from barman.process import ProcessManager
from barman.remote_status import RemoteStatusMixin
from barman.retention_policies import RetentionPolicyFactory
from barman.utils import human_readable_timedelta, pretty_size, timeout
from barman.wal_archiver import (FileWalArchiver, StreamingWalArchiver,
WalArchiver)
_logger = logging.getLogger(__name__)
class CheckStrategy(object):
"""
This strategy for the 'check' collects the results of
every check and does not print any message.
This basic class is also responsible for immediately
logging any performed check with an error in case of
check failure and a debug message in case of success.
"""
# create a namedtuple object called CheckResult to manage check results
CheckResult = namedtuple('CheckResult', 'server_name check status')
# Default list used as a filter to identify non-critical checks
NON_CRITICAL_CHECKS = ['minimum redundancy requirements',
'backup maximum age',
'failed backups',
'archiver errors']
def __init__(self, ignore_checks=NON_CRITICAL_CHECKS):
"""
Silent Strategy constructor
:param list ignore_checks: list of checks that can be ignored
"""
self.ignore_list = ignore_checks
self.check_result = []
self.has_error = False
def result(self, server_name, check, status, hint=None):
"""
Store the result of a check (with no output).
Log any check result (error or debug level).
:param str server_name: the server is being checked
:param str check: the check name
:param bool status: True if succeeded
:param str,None hint: hint to print if not None:
"""
if not status:
# If the name of the check is not in the filter list,
# treat it as a blocking error, then notify the error
# and change the status of the strategy
if check not in self.ignore_list:
self.has_error = True
_logger.error(
"Check '%s' failed for server '%s'" %
(check, server_name))
else:
# otherwise simply log the error (as info)
_logger.info(
"Ignoring failed check '%s' for server '%s'" %
(check, server_name))
else:
_logger.debug(
"Check '%s' succeeded for server '%s'" %
(check, server_name))
# Store the result and does not output anything
result = self.CheckResult(server_name, check, status)
self.check_result.append(result)
class CheckOutputStrategy(CheckStrategy):
"""
This strategy for the 'check' command immediately sends
the result of a check to the designated output channel.
This class derives from the basic CheckStrategy, reuses
the same logic and adds output messages.
"""
def __init__(self):
"""
Output Strategy constructor
"""
super(CheckOutputStrategy, self).__init__(ignore_checks=())
def result(self, server_name, check, status, hint=None):
"""
Output Strategy constructor
:param str server_name: the server being checked
:param str check: the check name
:param bool status: True if succeeded
:param str,None hint: hint to print if not None:
"""
# Call the basic method
super(CheckOutputStrategy, self).result(
server_name, check, status, hint)
# Send result to output
output.result('check', server_name, check, status, hint)
class Server(RemoteStatusMixin):
"""
This class represents the PostgreSQL server to backup.
"""
XLOG_DB = "xlog.db"
# the strategy for the management of the results of the various checks
__default_check_strategy = CheckOutputStrategy()
def __init__(self, config):
"""
Server constructor.
:param barman.config.ServerConfig config: the server configuration
"""
super(Server, self).__init__()
self.config = config
self.path = self._build_path(self.config.path_prefix)
self.process_manager = ProcessManager(self.config)
self.enforce_retention_policies = False
self.postgres = None
self.streaming = None
self.archivers = []
# Initialize the backup manager
self.backup_manager = BackupManager(self)
# Initialize the main PostgreSQL connection
try:
self.postgres = PostgreSQLConnection(config)
# If the PostgreSQLConnection creation fails, disable the Server
except ConninfoException as e:
self.config.disabled = True
self.config.msg_list.append("PostgreSQL connection: " +
str(e).strip())
# ARCHIVER_OFF_BACKCOMPATIBILITY - START OF CODE
# IMPORTANT: This is a back-compatibility feature that has
# been added in Barman 2.0. It highlights a deprecated
# behaviour, and helps users during this transition phase.
# It forces 'archiver=on' when both archiver and streaming_archiver
# are set to 'off' (default values) and displays a warning,
# requesting users to explicitly set the value in the configuration.
# When this back-compatibility feature will be removed from Barman
# (in a couple of major releases), developers will need to remove
# this block completely and reinstate the block of code you find
# a few lines below (search for ARCHIVER_OFF_BACKCOMPATIBILITY
# throughout the code).
if self.config.archiver is False and \
self.config.streaming_archiver is False:
output.warning("No archiver enabled for server '%s'. "
"Please turn on 'archiver', 'streaming_archiver' "
"or both", self.config.name)
output.warning("Forcing 'archiver = on'")
self.config.archiver = True
# ARCHIVER_OFF_BACKCOMPATIBILITY - END OF CODE
# Initialize the FileWalArchiver
# WARNING: Order of items in self.archivers list is important!
# The files will be archived in that order.
if self.config.archiver:
try:
self.archivers.append(FileWalArchiver(self.backup_manager))
except AttributeError as e:
_logger.debug(e)
self.config.disabled = True
self.config.msg_list.append('Unable to initialise the '
'file based archiver')
# Initialize the streaming PostgreSQL connection only when
# backup_method is postgres or the streaming_archiver is in use
if (self.config.backup_method == 'postgres' or
self.config.streaming_archiver):
try:
self.streaming = StreamingConnection(config)
# If the StreamingConnection creation fails, disable the server
except ConninfoException as e:
self.config.disabled = True
self.config.msg_list.append("Streaming connection: " +
str(e).strip())
# Initialize the StreamingWalArchiver
# WARNING: Order of items in self.archivers list is important!
# The files will be archived in that order.
if self.config.streaming_archiver:
try:
self.archivers.append(StreamingWalArchiver(
self.backup_manager))
# If the StreamingWalArchiver creation fails,
# disable the server
except AttributeError as e:
_logger.debug(e)
self.config.disabled = True
self.config.msg_list.append('Unable to initialise the '
'streaming archiver')
# IMPORTANT: The following lines of code have been
# temporarily commented in order to make the code
# back-compatible after the introduction of 'archiver=off'
# as default value in Barman 2.0.
# When the back compatibility feature for archiver will be
# removed, the following lines need to be decommented.
# ARCHIVER_OFF_BACKCOMPATIBILITY - START OF CODE
# # At least one of the available archive modes should be enabled
# if len(self.archivers) < 1:
# self.config.disabled = True
# self.config.msg_list.append(
# "No archiver enabled for server '%s'. "
# "Please turn on 'archiver', 'streaming_archiver' or both"
# % config.name)
# ARCHIVER_OFF_BACKCOMPATIBILITY - END OF CODE
# Sanity check: if file based archiver is disabled, and only
# WAL streaming is enabled, a replication slot name must be configured.
if not self.config.archiver and self.config.streaming_archiver and \
self.config.slot_name is None:
self.config.disabled = True
self.config.msg_list.append(
"Streaming-only archiver requires 'streaming_conninfo' and "
"'slot_name' options to be properly configured")
# Set bandwidth_limit
if self.config.bandwidth_limit:
try:
self.config.bandwidth_limit = int(self.config.bandwidth_limit)
except ValueError:
_logger.warning('Invalid bandwidth_limit "%s" for server "%s" '
'(fallback to "0")' % (
self.config.bandwidth_limit,
self.config.name))
self.config.bandwidth_limit = None
# set tablespace_bandwidth_limit
if self.config.tablespace_bandwidth_limit:
rules = {}
for rule in self.config.tablespace_bandwidth_limit.split():
try:
key, value = rule.split(':', 1)
value = int(value)
if value != self.config.bandwidth_limit:
rules[key] = value
except ValueError:
_logger.warning(
"Invalid tablespace_bandwidth_limit rule '%s'" % rule)
if len(rules) > 0:
self.config.tablespace_bandwidth_limit = rules
else:
self.config.tablespace_bandwidth_limit = None
# Set minimum redundancy (default 0)
if self.config.minimum_redundancy.isdigit():
self.config.minimum_redundancy = int(
self.config.minimum_redundancy)
if self.config.minimum_redundancy < 0:
_logger.warning('Negative value of minimum_redundancy "%s" '
'for server "%s" (fallback to "0")' % (
self.config.minimum_redundancy,
self.config.name))
self.config.minimum_redundancy = 0
else:
_logger.warning('Invalid minimum_redundancy "%s" for server "%s" '
'(fallback to "0")' % (
self.config.minimum_redundancy,
self.config.name))
self.config.minimum_redundancy = 0
# Initialise retention policies
self._init_retention_policies()
def _init_retention_policies(self):
# Set retention policy mode
if self.config.retention_policy_mode != 'auto':
_logger.warning(
'Unsupported retention_policy_mode "%s" for server "%s" '
'(fallback to "auto")' % (
self.config.retention_policy_mode, self.config.name))
self.config.retention_policy_mode = 'auto'
# If retention_policy is present, enforce them
if self.config.retention_policy:
# Check wal_retention_policy
if self.config.wal_retention_policy != 'main':
_logger.warning(
'Unsupported wal_retention_policy value "%s" '
'for server "%s" (fallback to "main")' % (
self.config.wal_retention_policy, self.config.name))
self.config.wal_retention_policy = 'main'
# Create retention policy objects
try:
rp = RetentionPolicyFactory.create(
self, 'retention_policy', self.config.retention_policy)
# Reassign the configuration value (we keep it in one place)
self.config.retention_policy = rp
_logger.debug('Retention policy for server %s: %s' % (
self.config.name, self.config.retention_policy))
try:
rp = RetentionPolicyFactory.create(
self, 'wal_retention_policy',
self.config.wal_retention_policy)
# Reassign the configuration value
# (we keep it in one place)
self.config.wal_retention_policy = rp
_logger.debug(
'WAL retention policy for server %s: %s' % (
self.config.name,
self.config.wal_retention_policy))
except ValueError:
_logger.exception(
'Invalid wal_retention_policy setting "%s" '
'for server "%s" (fallback to "main")' % (
self.config.wal_retention_policy,
self.config.name))
rp = RetentionPolicyFactory.create(
self, 'wal_retention_policy', 'main')
self.config.wal_retention_policy = rp
self.enforce_retention_policies = True
except ValueError:
_logger.exception(
'Invalid retention_policy setting "%s" for server "%s"' % (
self.config.retention_policy, self.config.name))
def close(self):
"""
Close all the open connections to PostgreSQL
"""
if self.postgres:
self.postgres.close()
if self.streaming:
self.streaming.close()
def check(self, check_strategy=__default_check_strategy):
"""
Implements the 'server check' command and makes sure SSH and PostgreSQL
connections work properly. It checks also that backup directories exist
(and if not, it creates them).
The check command will time out after a time interval defined by the
check_timeout configuration value (default 30 seconds)
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
try:
with timeout(self.config.check_timeout):
# Check WAL archive
self.check_archive(check_strategy)
# Check postgres configuration
self.check_postgres(check_strategy)
# Check barman directories from barman configuration
self.check_directories(check_strategy)
# Check retention policies
self.check_retention_policy_settings(check_strategy)
# Check for backup validity
self.check_backup_validity(check_strategy)
# Executes the backup manager set of checks
self.backup_manager.check(check_strategy)
# Check if the msg_list of the server
# contains messages and output eventual failures
self.check_configuration(check_strategy)
# Executes check() for every archiver, passing
# remote status information for efficiency
for archiver in self.archivers:
archiver.check(check_strategy)
# Check archiver errors
self.check_archiver_errors(check_strategy)
except TimeoutError:
# The check timed out.
# Add a failed entry to the check strategy for this.
check_strategy.result(self.config.name, 'check timeout', False,
'barman check command timed out')
def check_archive(self, check_strategy):
"""
Checks WAL archive
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
# Make sure that WAL archiving has been setup
# XLOG_DB needs to exist and its size must be > 0
# NOTE: we do not need to acquire a lock in this phase
xlogdb_empty = True
if os.path.exists(self.xlogdb_file_name):
with open(self.xlogdb_file_name, "rb") as fxlogdb:
if os.fstat(fxlogdb.fileno()).st_size > 0:
xlogdb_empty = False
# NOTE: This check needs to be only visible if it fails
if xlogdb_empty:
check_strategy.result(
self.config.name, 'WAL archive', False,
'please make sure WAL shipping is setup')
def check_postgres(self, check_strategy):
"""
Checks PostgreSQL connection
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
# Take the status of the remote server
remote_status = self.get_remote_status()
if remote_status.get('server_txt_version'):
check_strategy.result(self.config.name, 'PostgreSQL', True)
else:
check_strategy.result(self.config.name, 'PostgreSQL', False)
return
# Check for superuser privileges in PostgreSQL
if remote_status.get('is_superuser') is not None:
if remote_status.get('is_superuser'):
check_strategy.result(
self.config.name, 'superuser', True)
else:
check_strategy.result(
self.config.name, 'not superuser', False,
'superuser privileges for PostgreSQL connection required')
if 'streaming_supported' in remote_status:
hint = None
# If a streaming connection is available,
# add its status to the output of the check
if remote_status['streaming_supported'] is None:
hint = remote_status['connection_error']
elif not remote_status['streaming_supported']:
hint = ('Streaming connection not supported'
' for PostgreSQL < 9.2')
check_strategy.result(self.config.name, 'PostgreSQL streaming',
remote_status.get('streaming'), hint)
# Check wal_level parameter: must be different from 'minimal'
# the parameter has been introduced in postgres >= 9.0
if 'wal_level' in remote_status:
if remote_status['wal_level'] != 'minimal':
check_strategy.result(
self.config.name, 'wal_level', True)
else:
check_strategy.result(
self.config.name, 'wal_level', False,
"please set it to a higher level than 'minimal'")
# Check the presence and the status of the configured replication slot
# This check will be skipped if `slot_name` is undefined
if self.config.slot_name:
slot = remote_status['replication_slot']
# The streaming_archiver is enabled
if self.config.streaming_archiver is True:
# Error if PostgreSQL is too old
if not remote_status['replication_slot_support']:
check_strategy.result(
self.config.name,
'replication slot',
False,
"slot_name parameter set but "
"PostgreSQL server is too old (%s < 9.4)" % (
remote_status['server_txt_version']))
# Replication slots are supported
else:
# The slot is not present
if slot is None:
check_strategy.result(
self.config.name, 'replication slot', False,
"replication slot '%s' doesn't exist. "
"Please execute "
"'barman receive-wal --create-slot %s'" %
(self.config.slot_name, self.config.name))
else:
# The slot is present but not initialised
if slot.restart_lsn is None:
check_strategy.result(
self.config.name, 'replication slot', False,
"slot '%s' not initialised: "
"is 'receive-wal' running?" %
self.config.slot_name)
# The slot is present but not active
elif slot.active is False:
check_strategy.result(
self.config.name, 'replication slot', False,
"slot '%s' not active: "
"is 'receive-wal' running?" %
self.config.slot_name)
else:
check_strategy.result(self.config.name,
'replication slot',
True)
else:
# If the streaming_archiver is disabled and the slot_name
# option is present in the configuration, we check that
# a replication slot with the specified name is NOT present
# and NOT active.
# NOTE: This is not a failure, just a warning.
if slot is not None:
if slot.restart_lsn \
is not None:
slot_status = 'initialised'
# Check if the slot is also active
if slot.active:
slot_status = 'active'
# Warn the user
check_strategy.result(
self.config.name,
'replication slot',
True,
"WARNING: slot '%s' is %s but not required "
"by the current config" % (
self.config.slot_name, slot_status))
def _make_directories(self):
"""
Make backup directories in case they do not exist
"""
for key in self.config.KEYS:
if key.endswith('_directory') and hasattr(self.config, key):
val = getattr(self.config, key)
if val is not None and not os.path.isdir(val):
# noinspection PyTypeChecker
os.makedirs(val)
def check_directories(self, check_strategy):
"""
Checks backup directories and creates them if they do not exist
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
if not self.config.disabled:
try:
self._make_directories()
except OSError as e:
check_strategy.result(self.config.name, 'directories', False,
"%s: %s" % (e.filename, e.strerror))
else:
check_strategy.result(self.config.name, 'directories', True)
def check_configuration(self, check_strategy):
"""
Check for error messages in the message list
of the server and output eventual errors
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
if len(self.config.msg_list):
check_strategy.result(self.config.name, 'configuration', False)
for conflict_paths in self.config.msg_list:
output.info("\t\t%s" % conflict_paths)
def check_retention_policy_settings(self, check_strategy):
"""
Checks retention policy setting
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
if (self.config.retention_policy and
not self.enforce_retention_policies):
check_strategy.result(self.config.name,
'retention policy settings', False,
'see log')
else:
check_strategy.result(self.config.name,
'retention policy settings', True)
def check_backup_validity(self, check_strategy):
"""
Check if backup validity requirements are satisfied
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
# first check: check backup maximum age
if self.config.last_backup_maximum_age is not None:
# get maximum age information
backup_age = self.backup_manager.validate_last_backup_maximum_age(
self.config.last_backup_maximum_age)
# format the output
check_strategy.result(
self.config.name, 'backup maximum age',
backup_age[0],
"interval provided: %s, latest backup age: %s" % (
human_readable_timedelta(
self.config.last_backup_maximum_age), backup_age[1]))
else:
# last_backup_maximum_age provided by the user
check_strategy.result(
self.config.name, 'backup maximum age', True,
"no last_backup_maximum_age provided")
def check_archiver_errors(self, check_strategy):
"""
Checks the presence of archiving errors
:param CheckStrategy check_strategy: the strategy for the management
of the results of the check
"""
if os.path.isdir(self.config.errors_directory):
errors = os.listdir(self.config.errors_directory)
else:
errors = []
check_strategy.result(
self.config.name,
"archiver errors",
len(errors) == 0,
WalArchiver.summarise_error_files(errors)
)
def status_postgres(self):
"""
Status of PostgreSQL server
"""
remote_status = self.get_remote_status()
if remote_status['server_txt_version']:
output.result('status', self.config.name,
"pg_version",
"PostgreSQL version",
remote_status['server_txt_version'])
else:
output.result('status', self.config.name,
"pg_version",
"PostgreSQL version",
"FAILED trying to get PostgreSQL version")
return
if remote_status['pgespresso_installed']:
output.result('status', self.config.name, 'pgespresso',
'pgespresso extension', "Available")
else:
output.result('status', self.config.name, 'pgespresso',
'pgespresso extension', "Not available")
if remote_status.get('current_size') is not None:
output.result('status', self.config.name,
'current_size',
'Current data size',
pretty_size(remote_status['current_size']))
if remote_status['data_directory']:
output.result('status', self.config.name,
"data_directory",
"PostgreSQL Data directory",
remote_status['data_directory'])
if remote_status['current_xlog']:
output.result('status', self.config.name,
"current_xlog",
"Current WAL segment",
remote_status['current_xlog'])
def status_wal_archiver(self):
"""
Status of WAL archiver(s)
"""
for archiver in self.archivers:
archiver.status()
def status_retention_policies(self):
"""
Status of retention policies enforcement
"""
if self.enforce_retention_policies:
output.result('status', self.config.name,
"retention_policies",
"Retention policies",
"enforced "
"(mode: %s, retention: %s, WAL retention: %s)" % (
self.config.retention_policy_mode,
self.config.retention_policy,
self.config.wal_retention_policy))
else:
output.result('status', self.config.name,
"retention_policies",
"Retention policies",
"not enforced")
def status(self):
"""
Implements the 'server-status' command.
"""
if self.config.description:
output.result('status', self.config.name,
"description",
"Description", self.config.description)
output.result('status', self.config.name,
"active",
"Active", self.config.active)
output.result('status', self.config.name,
"disabled",
"Disabled", self.config.disabled)
self.status_postgres()
self.status_wal_archiver()
self.status_retention_policies()
# Executes the backup manager status info method
self.backup_manager.status()
def fetch_remote_status(self):
"""
Get the status of the remote server
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
result = {}
# Merge status for a postgres connection
if self.postgres:
result.update(self.postgres.get_remote_status())
# Merge status for a streaming connection
if self.streaming:
result.update(self.streaming.get_remote_status())
# Merge status for each archiver
for archiver in self.archivers:
result.update(archiver.get_remote_status())
# Merge status defined by the BackupManager
result.update(self.backup_manager.get_remote_status())
return result
def show(self):
"""
Shows the server configuration
"""
# Populate result map with all the required keys
result = self.config.to_json()
remote_status = self.get_remote_status()
result.update(remote_status)
# Backup maximum age section
if self.config.last_backup_maximum_age is not None:
age = self.backup_manager.validate_last_backup_maximum_age(
self.config.last_backup_maximum_age)
# If latest backup is between the limits of the
# last_backup_maximum_age configuration, display how old is
# the latest backup.
if age[0]:
msg = "%s (latest backup: %s )" % \
(human_readable_timedelta(
self.config.last_backup_maximum_age),
age[1])
else:
# If latest backup is outside the limits of the
# last_backup_maximum_age configuration (or the configuration
# value is none), warn the user.
msg = "%s (WARNING! latest backup is %s old)" % \
(human_readable_timedelta(
self.config.last_backup_maximum_age),
age[1])
result['last_backup_maximum_age'] = msg
else:
result['last_backup_maximum_age'] = "None"
output.result('show_server', self.config.name, result)
def delete_backup(self, backup):
"""Deletes a backup
:param backup: the backup to delete
"""
try:
# Lock acquisition: if you can acquire a ServerBackupLock
# it means that no backup process is running on that server,
# so there is no need to check the backup status.
# Simply proceed with the normal delete process.
server_backup_lock = ServerBackupLock(
self.config.barman_lock_directory,
self.config.name)
server_backup_lock.acquire(server_backup_lock.raise_if_fail,
server_backup_lock.wait)
server_backup_lock.release()
return self.backup_manager.delete_backup(backup)
except LockFileBusy:
# Otherwise if the lockfile is busy, a backup process is actually
# running on that server. To be sure that it's safe
# to delete the backup, we must check its status and its position
# in the catalogue.
# If it is the first and it is STARTED or EMPTY, we are trying to
# remove a running backup. This operation must be forbidden.
# Otherwise, normally delete the backup.
first_backup_id = self.get_first_backup_id(BackupInfo.STATUS_ALL)
if backup.backup_id == first_backup_id \
and backup.status in (BackupInfo.STARTED,
BackupInfo.EMPTY):
output.error("Cannot delete a running backup (%s %s)"
% (self.config.name, backup.backup_id))
else:
return self.backup_manager.delete_backup(backup)
except LockFilePermissionDenied as e:
# We cannot access the lockfile.
# Exit without removing the backup.
output.error("Permission denied, unable to access '%s'" % e)
def backup(self):
"""
Performs a backup for the server
"""
try:
# Default strategy for check in backup is CheckStrategy
# This strategy does not print any output - it only logs checks
strategy = CheckStrategy()
self.check(strategy)
if strategy.has_error:
output.error("Impossible to start the backup. Check the log "
"for more details, or run 'barman check %s'"
% self.config.name)
return
# check required backup directories exist
self._make_directories()
except OSError as e:
output.error('failed to create %s directory: %s',
e.filename, e.strerror)
return
# Make sure we are not wasting an precious streaming PostgreSQL
# connection that may have been opened by the self.check() call
if self.streaming:
self.streaming.close()
try:
# lock acquisition and backup execution
with ServerBackupLock(self.config.barman_lock_directory,
self.config.name):
self.backup_manager.backup()
# Archive incoming WALs and update WAL catalogue
self.archive_wal(verbose=False)
except LockFileBusy:
output.error("Another backup process is running")
except LockFilePermissionDenied as e:
output.error("Permission denied, unable to access '%s'" % e)
def get_available_backups(
self, status_filter=BackupManager.DEFAULT_STATUS_FILTER):
"""
Get a list of available backups
param: status_filter: the status of backups to return,
default to BackupManager.DEFAULT_STATUS_FILTER
"""
return self.backup_manager.get_available_backups(status_filter)
def get_last_backup_id(
self, status_filter=BackupManager.DEFAULT_STATUS_FILTER):
"""
Get the id of the latest/last backup in the catalog (if exists)
:param status_filter: The status of the backup to return,
default to DEFAULT_STATUS_FILTER.
:return string|None: ID of the backup
"""
return self.backup_manager.get_last_backup_id(status_filter)
def get_first_backup_id(
self, status_filter=BackupManager.DEFAULT_STATUS_FILTER):
"""
Get the id of the oldest/first backup in the catalog (if exists)
:param status_filter: The status of the backup to return,
default to DEFAULT_STATUS_FILTER.
:return string|None: ID of the backup
"""
return self.backup_manager.get_first_backup_id(status_filter)
def list_backups(self):
"""
Lists all the available backups for the server
"""
retention_status = self.report_backups()
backups = self.get_available_backups(BackupInfo.STATUS_ALL)
for key in sorted(backups.keys(), reverse=True):
backup = backups[key]
backup_size = backup.size or 0
wal_size = 0
rstatus = None
if backup.status == BackupInfo.DONE:
try:
wal_info = self.get_wal_info(backup)
backup_size += wal_info['wal_size']
wal_size = wal_info['wal_until_next_size']
except BadXlogSegmentName as e:
output.error(
"invalid xlog segment name %r\n"
"HINT: Please run \"barman rebuild-xlogdb %s\" "
"to solve this issue",
str(e), self.config.name)
if self.enforce_retention_policies and \
retention_status[backup.backup_id] != BackupInfo.VALID:
rstatus = retention_status[backup.backup_id]
output.result('list_backup', backup, backup_size, wal_size,
rstatus)
def get_backup(self, backup_id):
"""
Return the backup information for the given backup id.
If the backup_id is None or backup.info file doesn't exists,
it returns None.
:param str|None backup_id: the ID of the backup to return
:rtype: BackupInfo|None
"""
return self.backup_manager.get_backup(backup_id)
def get_previous_backup(self, backup_id):
"""
Get the previous backup (if any) from the catalog
:param backup_id: the backup id from which return the previous
"""
return self.backup_manager.get_previous_backup(backup_id)
def get_next_backup(self, backup_id):
"""
Get the next backup (if any) from the catalog
:param backup_id: the backup id from which return the next
"""
return self.backup_manager.get_next_backup(backup_id)
def get_required_xlog_files(self, backup, target_tli=None,
target_time=None, target_xid=None):
"""
Get the xlog files required for a recovery
"""
begin = backup.begin_wal
end = backup.end_wal
# If timeline isn't specified, assume it is the same timeline
# of the backup
if not target_tli:
target_tli, _, _ = xlog.decode_segment_name(end)
with self.xlogdb() as fxlogdb:
for line in fxlogdb:
wal_info = WalFileInfo.from_xlogdb_line(line)
# Handle .history files: add all of them to the output,
# regardless of their age
if xlog.is_history_file(wal_info.name):
yield wal_info
continue
if wal_info.name < begin:
continue
tli, _, _ = xlog.decode_segment_name(wal_info.name)
if tli > target_tli:
continue
yield wal_info
if wal_info.name > end:
end = wal_info.name
if target_time and target_time < wal_info.time:
break
# return all the remaining history files
for line in fxlogdb:
wal_info = WalFileInfo.from_xlogdb_line(line)
if xlog.is_history_file(wal_info.name):
yield wal_info
# TODO: merge with the previous
def get_wal_until_next_backup(self, backup, include_history=False):
"""
Get the xlog files between backup and the next
:param BackupInfo backup: a backup object, the starting point
to retrieve WALs
:param bool include_history: option for the inclusion of
include_history files into the output
"""
begin = backup.begin_wal
next_end = None
if self.get_next_backup(backup.backup_id):
next_end = self.get_next_backup(backup.backup_id).end_wal
backup_tli, _, _ = xlog.decode_segment_name(begin)
with self.xlogdb() as fxlogdb:
for line in fxlogdb:
wal_info = WalFileInfo.from_xlogdb_line(line)
# Handle .history files: add all of them to the output,
# regardless of their age, if requested (the 'include_history'
# parameter is True)
if xlog.is_history_file(wal_info.name):
if include_history:
yield wal_info
continue
if wal_info.name < begin:
continue
tli, _, _ = xlog.decode_segment_name(wal_info.name)
if tli > backup_tli:
continue
if not xlog.is_wal_file(wal_info.name):
continue
if next_end and wal_info.name > next_end:
break
yield wal_info
def get_wal_full_path(self, wal_name):
"""
Build the full path of a WAL for a server given the name
:param wal_name: WAL file name
"""
# Build the path which contains the file
hash_dir = os.path.join(self.config.wals_directory,
xlog.hash_dir(wal_name))
# Build the WAL file full path
full_path = os.path.join(hash_dir, wal_name)
return full_path
def get_wal_info(self, backup_info):
"""
Returns information about WALs for the given backup
:param BackupInfo backup_info: the target backup
"""
begin = backup_info.begin_wal
end = backup_info.end_wal
# counters
wal_info = dict.fromkeys(
('wal_num', 'wal_size',
'wal_until_next_num', 'wal_until_next_size',
'wal_until_next_compression_ratio',
'wal_compression_ratio'), 0)
# First WAL (always equal to begin_wal) and Last WAL names and ts
wal_info['wal_first'] = None
wal_info['wal_first_timestamp'] = None
wal_info['wal_last'] = None
wal_info['wal_last_timestamp'] = None
# WAL rate (default 0.0 per second)
wal_info['wals_per_second'] = 0.0
for item in self.get_wal_until_next_backup(backup_info):
if item.name == begin:
wal_info['wal_first'] = item.name
wal_info['wal_first_timestamp'] = item.time
if item.name <= end:
wal_info['wal_num'] += 1
wal_info['wal_size'] += item.size
else:
wal_info['wal_until_next_num'] += 1
wal_info['wal_until_next_size'] += item.size
wal_info['wal_last'] = item.name
wal_info['wal_last_timestamp'] = item.time
# Calculate statistics only for complete backups
# If the cron is not running for any reason, the required
# WAL files could be missing
if wal_info['wal_first'] and wal_info['wal_last']:
# Estimate WAL ratio
# Calculate the difference between the timestamps of
# the first WAL (begin of backup) and the last WAL
# associated to the current backup
wal_info['wal_total_seconds'] = (
wal_info['wal_last_timestamp'] -
wal_info['wal_first_timestamp'])
if wal_info['wal_total_seconds'] > 0:
wal_info['wals_per_second'] = (
float(wal_info['wal_num'] +
wal_info['wal_until_next_num']) /
wal_info['wal_total_seconds'])
# evaluation of compression ratio for basebackup WAL files
wal_info['wal_theoretical_size'] = \
wal_info['wal_num'] * float(xlog.XLOG_SEG_SIZE)
try:
wal_info['wal_compression_ratio'] = 1 - (
wal_info['wal_size'] /
wal_info['wal_theoretical_size'])
except ZeroDivisionError:
wal_info['wal_compression_ratio'] = 0.0
# evaluation of compression ratio of WAL files
wal_info['wal_until_next_theoretical_size'] = \
wal_info['wal_until_next_num'] * float(xlog.XLOG_SEG_SIZE)
try:
wal_info['wal_until_next_compression_ratio'] = 1 - (
wal_info['wal_until_next_size'] /
wal_info['wal_until_next_theoretical_size'])
except ZeroDivisionError:
wal_info['wal_until_next_compression_ratio'] = 0.0
return wal_info
def recover(self, backup_info, dest, tablespaces=None, target_tli=None,
target_time=None, target_xid=None, target_name=None,
exclusive=False, remote_command=None):
"""
Performs a recovery of a backup
:param barman.infofile.BackupInfo backup_info: the backup to recover
:param str dest: the destination directory
:param dict[str,str]|None tablespaces: a tablespace
name -> location map (for relocation)
:param str|None target_tli: the target timeline
:param str|None target_time: the target time
:param str|None target_xid: the target xid
:param str|None target_name: the target name created previously with
pg_create_restore_point() function call
:param bool exclusive: whether the recovery is exclusive or not
:param str|None remote_command: default None. The remote command to
recover the base backup, in case of remote backup.
"""
return self.backup_manager.recover(
backup_info, dest, tablespaces, target_tli, target_time,
target_xid, target_name, exclusive, remote_command)
def get_wal(self, wal_name, compression=None, output_directory=None,
peek=None):
"""
Retrieve a WAL file from the archive
:param str wal_name: id of the WAL file to find into the WAL archive
:param str|None compression: compression format for the output
:param str|None output_directory: directory where to deposit the
WAL file
:param int|None peek: if defined list the next N WAL file
"""
# Sanity check
if not xlog.is_any_xlog_file(wal_name):
output.error("'%s' is not a valid wal file name", wal_name)
return
# If peek is requested we only output a list of files
if peek:
# Get the next ``peek`` files following the provided ``wal_name``.
# If ``wal_name`` is not a simple wal file,
# we cannot guess the names of the following WAL files.
# So ``wal_name`` is the only possible result, if exists.
if xlog.is_wal_file(wal_name):
wal_peek_list = itertools.islice(
xlog.generate_segment_names(wal_name), peek)
else:
wal_peek_list = [wal_name]
# Output the content of wal_peek_list until we find a missing file
for wal_peek_name in wal_peek_list:
wal_peek_file = self.get_wal_full_path(wal_peek_name)
# If ``wal_peek_file`` doesn't exist, stop the process
if not os.path.exists(wal_peek_file):
return
output.info(wal_peek_name, log=False)
# Do not output anything else
return
# Get the WAL file full path
wal_file = self.get_wal_full_path(wal_name)
# Check for file existence
if not os.path.exists(wal_file):
output.error("WAL file '%s' not found in server '%s'",
wal_name, self.config.name)
return
# If an output directory was provided write the file inside it
# otherwise we use standard output
if output_directory is not None:
destination_path = os.path.join(output_directory, wal_name)
try:
destination = open(destination_path, 'w')
output.info("Writing WAL '%s' for server '%s' into '%s' file",
wal_name, self.config.name, destination_path)
except IOError as e:
output.error("Unable to open '%s' file: %s" %
destination_path, e)
return
else:
destination = sys.stdout
# Get a decompressor for the file (None if not compressed)
wal_compressor = self.backup_manager.compression_manager \
.get_compressor(compression=identify_compression(wal_file))
# Get a compressor for the output (None if not compressed)
# Here we need to handle explicitly the None value because we don't
# want it ot fallback to the configured compression
if compression is not None:
out_compressor = self.backup_manager.compression_manager \
.get_compressor(compression=compression)
else:
out_compressor = None
# Initially our source is the stored WAL file and we do not have
# any temporary file
source_file = wal_file
uncompressed_file = None
compressed_file = None
# If the required compression is different from the source we
# decompress/compress it into the required format (getattr is
# used here to gracefully handle None objects)
if getattr(wal_compressor, 'compression', None) != \
getattr(out_compressor, 'compression', None):
# If source is compressed, decompress it into a temporary file
if wal_compressor is not None:
uncompressed_file = NamedTemporaryFile(
dir=self.config.wals_directory,
prefix='.%s.' % wal_name,
suffix='.uncompressed')
# decompress wal file
wal_compressor.decompress(source_file, uncompressed_file.name)
source_file = uncompressed_file.name
# If output compression is required compress the source
# into a temporary file
if out_compressor is not None:
compressed_file = NamedTemporaryFile(
dir=self.config.wals_directory,
prefix='.%s.' % wal_name,
suffix='.compressed')
out_compressor.compress(source_file, compressed_file.name)
source_file = compressed_file.name
# Copy the prepared source file to destination
with open(source_file) as input_file:
shutil.copyfileobj(input_file, destination)
# Remove temp files
if uncompressed_file is not None:
uncompressed_file.close()
if compressed_file is not None:
compressed_file.close()
def cron(self, wals=True, retention_policies=True):
"""
Maintenance operations
:param bool wals: WAL archive maintenance
:param bool retention_policies: retention policy maintenance
"""
try:
# Actually this is the highest level of locking in the cron,
# this stops the execution of multiple cron on the same server
with ServerCronLock(self.config.barman_lock_directory,
self.config.name):
# WAL management and maintenance
if wals:
# Execute the archive-wal sub-process
self.cron_archive_wal()
if self.config.streaming_archiver:
# Spawn the receive-wal sub-process
self.cron_receive_wal()
else:
# Terminate the receive-wal sub-process if present
self.kill('receive-wal', fail_if_not_present=False)
# Retention policies execution
if retention_policies:
self.backup_manager.cron_retention_policy()
except LockFileBusy:
output.info(
"Another cron process is already running on server %s. "
"Skipping to the next server" % self.config.name)
except LockFilePermissionDenied as e:
output.error("Permission denied, unable to access '%s'" % e)
except (OSError, IOError) as e:
output.error("%s", e)
def cron_archive_wal(self):
"""
Method that handles the start of an 'archive-wal' sub-process.
This method must be run protected by ServerCronLock
"""
try:
# Try to acquire ServerWalArchiveLock, if the lock is available,
# no other 'archive-wal' processes are running on this server.
#
# There is a very little race condition window here because
# even if we are protected by ServerCronLock, the user could run
# another 'archive-wal' command manually. However, it would result
# in one of the two commands failing on lock acquisition,
# with no other consequence.
with ServerWalArchiveLock(
self.config.barman_lock_directory,
self.config.name):
# Output and release the lock immediately
output.info("Starting WAL archiving for server %s",
self.config.name, log=False)
# Init a Barman sub-process object
archive_process = BarmanSubProcess(
subcommand='archive-wal',
config=barman.__config__.config_file,
args=[self.config.name])
# Launch the sub-process
archive_process.execute()
except LockFileBusy:
# Another archive process is running for the server,
# warn the user and skip to the next sever.
output.info(
"Another archive-wal process is already running "
"on server %s. Skipping to the next server"
% self.config.name)
def cron_receive_wal(self):
"""
Method that handles the start of a 'receive-wal' sub process
This method must be run protected by ServerCronLock
"""
try:
# Try to acquire ServerWalReceiveLock, if the lock is available,
# no other 'receive-wal' processes are running on this server.
#
# There is a very little race condition window here because
# even if we are protected by ServerCronLock, the user could run
# another 'receive-wal' command manually. However, it would result
# in one of the two commands failing on lock acquisition,
# with no other consequence.
with ServerWalReceiveLock(
self.config.barman_lock_directory,
self.config.name):
# Output and release the lock immediately
output.info("Starting streaming archiver "
"for server %s",
self.config.name, log=False)
# Start a new receive-wal process
receive_process = BarmanSubProcess(
subcommand='receive-wal',
config=barman.__config__.config_file,
args=[self.config.name])
# Launch the sub-process
receive_process.execute()
except LockFileBusy:
# Another receive-wal process is running for the server
# exit without message
_logger.debug("Another STREAMING ARCHIVER process is running for "
"server %s" % self.config.name)
def archive_wal(self, verbose=True):
"""
Perform the WAL archiving operations.
Usually run as subprocess of the barman cron command,
but can be executed manually using the barman archive-wal command
:param bool verbose: if false outputs something only if there is
at least one file
"""
output.debug("Starting archive-wal for server %s", self.config.name)
try:
# Take care of the archive lock.
# Only one archive job per server is admitted
with ServerWalArchiveLock(self.config.barman_lock_directory,
self.config.name):
self.backup_manager.archive_wal(verbose)
except LockFileBusy:
# If another process is running for this server,
# warn the user and skip to the next server
output.info("Another archive-wal process is already running "
"on server %s. Skipping to the next server"
% self.config.name)
def create_physical_repslot(self):
"""
Create a physical replication slot using the streaming connection
"""
if not self.streaming:
output.error("Unable to create a physical replication slot: "
"streaming connection not configured")
return
# Replication slots are not supported by PostgreSQL < 9.4
try:
if self.streaming.server_version < 90400:
output.error("Unable to create a physical replication slot: "
"not supported by '%s' "
"(9.4 or higher is required)" %
self.streaming.server_major_version)
return
except PostgresException as exc:
msg = "Cannot connect to server '%s'" % self.config.name
output.error(msg, log=False)
_logger.error("%s: %s", msg, str(exc).strip())
return
if not self.config.slot_name:
output.error("Unable to create a physical replication slot: "
"slot_name configuration option required")
return
output.info(
"Creating physical replication slot '%s' on server '%s'",
self.config.slot_name,
self.config.name)
try:
self.streaming.create_physical_repslot(self.config.slot_name)
output.info("Replication slot '%s' created", self.config.slot_name)
except PostgresDuplicateReplicationSlot:
output.error("Replication slot '%s' already exists",
self.config.slot_name)
except PostgresReplicationSlotsFull:
output.error("All replication slots for server '%s' are in use\n"
"Free one or increase the max_replication_slots "
"value on your PostgreSQL server.",
self.config.name)
except PostgresException as exc:
output.error(
"Cannot create replication slot '%s' on server '%s': %s",
self.config.slot_name,
self.config.name,
str(exc).strip())
def drop_repslot(self):
"""
Drop a replication slot using the streaming connection
"""
if not self.streaming:
output.error("Unable to drop a physical replication slot: "
"streaming connection not configured")
return
# Replication slots are not supported by PostgreSQL < 9.4
try:
if self.streaming.server_version < 90400:
output.error("Unable to drop a physical replication slot: "
"not supported by '%s' (9.4 or higher is "
"required)" %
self.streaming.server_major_version)
return
except PostgresException as exc:
msg = "Cannot connect to server '%s'" % self.config.name
output.error(msg, log=False)
_logger.error("%s: %s", msg, str(exc).strip())
return
if not self.config.slot_name:
output.error("Unable to drop a physical replication slot: "
"slot_name configuration option required")
return
output.info(
"Dropping physical replication slot '%s' on server '%s'",
self.config.slot_name,
self.config.name)
try:
self.streaming.drop_repslot(self.config.slot_name)
output.info("Replication slot '%s' dropped", self.config.slot_name)
except PostgresInvalidReplicationSlot:
output.error("Replication slot '%s' does not exist",
self.config.slot_name)
except PostgresReplicationSlotInUse as exc:
output.error(
"Cannot drop replication slot '%s' on server '%s' "
"because it is in use.",
self.config.slot_name,
self.config.name)
except PostgresException as exc:
output.error(
"Cannot drop replication slot '%s' on server '%s': %s",
self.config.slot_name,
self.config.name,
str(exc).strip())
def receive_wal(self, reset=False):
"""
Enable the reception of WAL files using streaming protocol.
Usually started by barman cron command.
Executing this manually, the barman process will not terminate but
will continuously receive WAL files from the PostgreSQL server.
:param reset: When set, resets the status of receive-wal
"""
# Execute the receive-wal command only if streaming_archiver
# is enabled
if not self.config.streaming_archiver:
output.error("Unable to start receive-wal process: "
"streaming_archiver option set to 'off' in "
"barman configuration file")
return
output.info("Starting receive-wal for server %s", self.config.name)
try:
# Take care of the receive-wal lock.
# Only one receiving process per server is permitted
with ServerWalReceiveLock(self.config.barman_lock_directory,
self.config.name):
try:
# Only the StreamingWalArchiver implementation
# does something.
# WARNING: This codes assumes that there is only one
# StreamingWalArchiver in the archivers list.
for archiver in self.archivers:
archiver.receive_wal(reset)
except ArchiverFailure as e:
output.error(e)
except LockFileBusy:
# If another process is running for this server,
if reset:
output.info("Unable to reset the status of receive-wal "
"for server %s. Process is still running"
% self.config.name)
else:
output.info("Another receive-wal process is already running "
"for server %s." % self.config.name)
@property
def xlogdb_file_name(self):
"""
The name of the file containing the XLOG_DB
:return str: the name of the file that contains the XLOG_DB
"""
return os.path.join(self.config.wals_directory, self.XLOG_DB)
@contextmanager
def xlogdb(self, mode='r'):
"""
Context manager to access the xlogdb file.
This method uses locking to make sure only one process is accessing
the database at a time. The database file will be created
if it not exists.
Usage example:
with server.xlogdb('w') as file:
file.write(new_line)
:param str mode: open the file with the required mode
(default read-only)
"""
if not os.path.exists(self.config.wals_directory):
os.makedirs(self.config.wals_directory)
xlogdb = self.xlogdb_file_name
with ServerXLOGDBLock(self.config.barman_lock_directory,
self.config.name):
# If the file doesn't exist and it is required to read it,
# we open it in a+ mode, to be sure it will be created
if not os.path.exists(xlogdb) and mode.startswith('r'):
if '+' not in mode:
mode = "a%s+" % mode[1:]
else:
mode = "a%s" % mode[1:]
with open(xlogdb, mode) as f:
# execute the block nested in the with statement
try:
yield f
finally:
# we are exiting the context
# if file is writable (mode contains w, a or +)
# make sure the data is written to disk
# http://docs.python.org/2/library/os.html#os.fsync
if any((c in 'wa+') for c in f.mode):
f.flush()
os.fsync(f.fileno())
def report_backups(self):
if not self.enforce_retention_policies:
return dict()
else:
return self.config.retention_policy.report()
def rebuild_xlogdb(self):
"""
Rebuild the whole xlog database guessing it from the archive content.
"""
return self.backup_manager.rebuild_xlogdb()
def get_backup_ext_info(self, backup_info):
"""
Return a dictionary containing all available information about a backup
The result is equivalent to the sum of information from
* BackupInfo object
* the Server.get_wal_info() return value
* the context in the catalog (if available)
* the retention policy status
:param backup_info: the target backup
:rtype dict: all information about a backup
"""
backup_ext_info = backup_info.to_dict()
if backup_info.status == BackupInfo.DONE:
try:
previous_backup = self.backup_manager.get_previous_backup(
backup_ext_info['backup_id'])
next_backup = self.backup_manager.get_next_backup(
backup_ext_info['backup_id'])
if previous_backup:
backup_ext_info[
'previous_backup_id'] = previous_backup.backup_id
else:
backup_ext_info['previous_backup_id'] = None
if next_backup:
backup_ext_info['next_backup_id'] = next_backup.backup_id
else:
backup_ext_info['next_backup_id'] = None
except UnknownBackupIdException:
# no next_backup_id and previous_backup_id items
# means "Not available"
pass
backup_ext_info.update(self.get_wal_info(backup_info))
if self.enforce_retention_policies:
policy = self.config.retention_policy
backup_ext_info['retention_policy_status'] = \
policy.backup_status(backup_info.backup_id)
else:
backup_ext_info['retention_policy_status'] = None
# Check any child timeline exists
children_timelines = self.get_children_timelines(
backup_ext_info['timeline'],
forked_after=backup_info.end_xlog)
backup_ext_info['children_timelines'] = \
children_timelines
return backup_ext_info
def show_backup(self, backup_info):
"""
Output all available information about a backup
:param backup_info: the target backup
"""
try:
backup_ext_info = self.get_backup_ext_info(backup_info)
output.result('show_backup', backup_ext_info)
except BadXlogSegmentName as e:
output.error(
"invalid xlog segment name %r\n"
"HINT: Please run \"barman rebuild-xlogdb %s\" "
"to solve this issue" %
str(e), self.config.name)
output.close_and_exit()
@staticmethod
def _build_path(path_prefix=None):
"""
If a path_prefix is provided build a string suitable to be used in
PATH environment variable by joining the path_prefix with the
current content of PATH environment variable.
If the `path_prefix` is None returns None.
:rtype: str|None
"""
if not path_prefix:
return None
sys_path = os.environ.get('PATH')
return "%s%s%s" % (path_prefix, os.pathsep, sys_path)
def kill(self, task, fail_if_not_present=True):
"""
Given the name of a barman sub-task type,
attempts to stop all the processes
:param string task: The task we want to stop
:param bool fail_if_not_present: Display an error when the process
is not present (default: True)
"""
process_list = self.process_manager.list(task)
for process in process_list:
if self.process_manager.kill(process):
output.info('Stopped process %s(%s)',
process.task, process.pid)
return
else:
output.error('Cannot terminate process %s(%s)',
process.task, process.pid)
return
if fail_if_not_present:
output.error('Termination of %s failed: '
'no such process for server %s',
task, self.config.name)
def switch_xlog(self, force=False, archive=None, archive_timeout=None):
"""
Execute the switch-xlog command on the target server
"""
try:
if force:
# If called with force, execute a checkpoint before the
# switch_xlog command
_logger.info('Force a CHECKPOINT before pg_switch_xlog()')
self.postgres.checkpoint()
# Perform the switch_xlog. expect a WAL name only if the switch
# has been successfully executed, False otherwise.
closed_xlog = self.postgres.switch_xlog()
if closed_xlog is None:
# Something went wrong during the execution of the
# pg_switch_xlog command
output.error("Unable to perform pg_switch_xlog "
"for server '%s'." % self.config.name)
return
if closed_xlog:
# The switch_xlog command have been executed successfully
output.info(
"The xlog file %s has been closed on server '%s'" %
(closed_xlog, self.config.name))
# If the user has asked to wait for a WAL file to be archived,
# wait until a new WAL file has been found
# or the timeout has expired
if archive:
output.info(
"Waiting for the xlog file %s from server '%s' "
"(max: %s seconds)",
closed_xlog, self.config.name, archive_timeout)
# Wait for a new file until end_time
end_time = time.time() + archive_timeout
while time.time() < end_time:
self.backup_manager.archive_wal(verbose=False)
# Finish if the closed wal file is in the archive.
if os.path.exists(self.get_wal_full_path(closed_xlog)):
break
# sleep a bit before retrying
time.sleep(.1)
else:
output.error("The xlog file %s has not been received "
"in %s seconds",
closed_xlog, archive_timeout)
else:
# Is not necessary to perform a switch_xlog
output.info("No switch required for server '%s'" %
self.config.name)
except PostgresIsInRecovery:
output.info("No switch performed because server '%s' "
"is a standby." % self.config.name)
except PostgresSuperuserRequired:
# Superuser rights are required to perform the switch_xlog
output.error("Barman switch-xlog requires superuser rights")
def replication_status(self, target='all'):
"""
Implements the 'replication-status' command.
"""
if target == 'hot-standby':
client_type = PostgreSQLConnection.STANDBY
elif target == 'wal-streamer':
client_type = PostgreSQLConnection.WALSTREAMER
else:
client_type = PostgreSQLConnection.ANY_STREAMING_CLIENT
try:
standby_info = self.postgres.get_replication_stats(client_type)
if standby_info is None:
output.error('Unable to connect to server %s' %
self.config.name)
else:
output.result('replication_status', self.config.name,
target, self.postgres.current_xlog_location,
standby_info)
except PostgresUnsupportedFeature as e:
output.info(" Requires PostgreSQL %s or higher", e)
except PostgresSuperuserRequired:
output.info(" Requires superuser rights")
def get_children_timelines(self, tli, forked_after=None):
"""
Get a list of the children of the passed timeline
:param int tli: Id of the timeline to check
:param str forked_after: XLog location after which the timeline
must have been created
:return List[xlog.HistoryFileData]: the list of timelines that
have the timeline with id 'tli' as parent
"""
if forked_after:
forked_after = xlog.parse_lsn(forked_after)
children = []
# Search all the history files after the passed timeline
children_tli = tli
while True:
children_tli += 1
history_path = os.path.join(self.config.wals_directory,
"%08X.history" % children_tli)
# If the file doesn't exists, stop searching
if not os.path.exists(history_path):
break
# Create the WalFileInfo object using the file
wal_info = WalFileInfo.from_file(history_path)
# Get content of the file. We need to pass a compressor manager
# here to handle an eventual compression of the history file
history_info = xlog.decode_history_file(
wal_info,
self.backup_manager.compression_manager)
# Save the history only if is reachable from this timeline.
for tinfo in history_info:
# The history file contains the full genealogy
# but we keep only the line with `tli` timeline as parent.
if tinfo.parent_tli != tli:
continue
# We need to return this history info only if this timeline
# has been forked after the passed LSN
if forked_after and tinfo.switchpoint < forked_after:
continue
children.append(tinfo)
return children
barman-2.1/barman/output.py 0000664 0001750 0001750 00000111611 13015546105 016365 0 ustar gcalacoci gcalacoci # Copyright (C) 2013-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module control how the output of Barman will be rendered
"""
from __future__ import print_function
import inspect
import logging
import sys
from barman.infofile import BackupInfo
from barman.utils import pretty_size
from barman.xlog import diff_lsn
__all__ = [
'error_occurred', 'debug', 'info', 'warning', 'error', 'exception',
'result', 'close_and_exit', 'close', 'set_output_writer',
'AVAILABLE_WRITERS', 'DEFAULT_WRITER', 'ConsoleOutputWriter',
'NagiosOutputWriter',
]
#: True if error or exception methods have been called
error_occurred = False
#: Exit code if error occurred
error_exit_code = 1
def _format_message(message, args):
"""
Format a message using the args list. The result will be equivalent to
message % args
If args list contains a dictionary as its only element the result will be
message % args[0]
:param str message: the template string to be formatted
:param tuple args: a list of arguments
:return: the formatted message
:rtype: str
"""
if len(args) == 1 and isinstance(args[0], dict):
return message % args[0]
elif len(args) > 0:
return message % args
else:
return message
def _put(level, message, *args, **kwargs):
"""
Send the message with all the remaining positional arguments to
the configured output manager with the right output level. The message will
be sent also to the logger unless explicitly disabled with log=False
No checks are performed on level parameter as this method is meant
to be called only by this module.
If level == 'exception' the stack trace will be also logged
:param str level:
:param str message: the template string to be formatted
:param tuple args: all remaining arguments are passed to the log formatter
:key bool log: whether to log the message
:key bool is_error: treat this message as an error
"""
# handle keyword-only parameters
log = kwargs.pop('log', True)
is_error = kwargs.pop('is_error', False)
if len(kwargs):
raise TypeError('%s() got an unexpected keyword argument %r'
% (inspect.stack()[1][3], kwargs.popitem()[0]))
if is_error:
global error_occurred
error_occurred = True
_writer.error_occurred()
# dispatch the call to the output handler
getattr(_writer, level)(message, *args)
# log the message as originating from caller's caller module
if log:
exc_info = False
if level == 'exception':
level = 'error'
exc_info = True
frm = inspect.stack()[2]
mod = inspect.getmodule(frm[0])
logger = logging.getLogger(mod.__name__)
log_level = logging.getLevelName(level.upper())
logger.log(log_level, message, *args, **{'exc_info': exc_info})
def _dispatch(obj, prefix, name, *args, **kwargs):
"""
Dispatch the call to the %(prefix)s_%(name) method of the obj object
:param obj: the target object
:param str prefix: prefix of the method to be called
:param str name: name of the method to be called
:param tuple args: all remaining positional arguments will be sent
to target
:param dict kwargs: all remaining keyword arguments will be sent to target
:return: the result of the invoked method
:raise ValueError: if the target method is not present
"""
method_name = "%s_%s" % (prefix, name)
handler = getattr(obj, method_name, None)
if callable(handler):
return handler(*args, **kwargs)
else:
raise ValueError("The object %r does not have the %r method" % (
obj, method_name))
def is_quiet():
"""
Calls the "is_quiet" method, accessing the protected parameter _quiet
of the instanced OutputWriter
:return bool: the _quiet parameter value
"""
return _writer.is_quiet()
def is_debug():
"""
Calls the "is_debug" method, accessing the protected parameter _debug
of the instanced OutputWriter
:return bool: the _debug parameter value
"""
return _writer.is_debug()
def debug(message, *args, **kwargs):
"""
Output a message with severity 'DEBUG'
:key bool log: whether to log the message
"""
_put('debug', message, *args, **kwargs)
def info(message, *args, **kwargs):
"""
Output a message with severity 'INFO'
:key bool log: whether to log the message
"""
_put('info', message, *args, **kwargs)
def warning(message, *args, **kwargs):
"""
Output a message with severity 'INFO'
:key bool log: whether to log the message
"""
_put('warning', message, *args, **kwargs)
def error(message, *args, **kwargs):
"""
Output a message with severity 'ERROR'.
Also records that an error has occurred unless the ignore parameter
is True.
:key bool ignore: avoid setting an error exit status (default False)
:key bool log: whether to log the message
"""
# ignore is a keyword-only parameter
ignore = kwargs.pop('ignore', False)
if not ignore:
kwargs.setdefault('is_error', True)
_put('error', message, *args, **kwargs)
def exception(message, *args, **kwargs):
"""
Output a message with severity 'EXCEPTION'
If raise_exception parameter doesn't evaluate to false raise and exception:
- if raise_exception is callable raise the result of raise_exception()
- if raise_exception is an exception raise it
- else raise the last exception again
:key bool ignore: avoid setting an error exit status
:key raise_exception:
raise an exception after the message has been processed
:key bool log: whether to log the message
"""
# ignore and raise_exception are keyword-only parameters
ignore = kwargs.pop('ignore', False)
# noinspection PyNoneFunctionAssignment
raise_exception = kwargs.pop('raise_exception', None)
if not ignore:
kwargs.setdefault('is_error', True)
_put('exception', message, *args, **kwargs)
if raise_exception:
if callable(raise_exception):
# noinspection PyCallingNonCallable
raise raise_exception(message)
elif isinstance(raise_exception, BaseException):
raise raise_exception
else:
raise
def init(command, *args, **kwargs):
"""
Initialize the output writer for a given command.
:param str command: name of the command are being executed
:param tuple args: all remaining positional arguments will be sent
to the output processor
:param dict kwargs: all keyword arguments will be sent
to the output processor
"""
try:
_dispatch(_writer, 'init', command, *args, **kwargs)
except ValueError:
exception('The %s writer does not support the "%s" command',
_writer.__class__.__name__, command)
close_and_exit()
def result(command, *args, **kwargs):
"""
Output the result of an operation.
:param str command: name of the command are being executed
:param tuple args: all remaining positional arguments will be sent
to the output processor
:param dict kwargs: all keyword arguments will be sent
to the output processor
"""
try:
_dispatch(_writer, 'result', command, *args, **kwargs)
except ValueError:
exception('The %s writer does not support the "%s" command',
_writer.__class__.__name__, command)
close_and_exit()
def close_and_exit():
"""
Close the output writer and terminate the program.
If an error has been emitted the program will report a non zero return
value.
"""
close()
if error_occurred:
sys.exit(error_exit_code)
else:
sys.exit(0)
def close():
"""
Close the output writer.
"""
_writer.close()
def set_output_writer(new_writer, *args, **kwargs):
"""
Replace the current output writer with a new one.
The new_writer parameter can be a symbolic name or an OutputWriter object
:param new_writer: the OutputWriter name or the actual OutputWriter
:type: string or an OutputWriter
:param tuple args: all remaining positional arguments will be passed
to the OutputWriter constructor
:param dict kwargs: all remaining keyword arguments will be passed
to the OutputWriter constructor
"""
global _writer
_writer.close()
if new_writer in AVAILABLE_WRITERS:
_writer = AVAILABLE_WRITERS[new_writer](*args, **kwargs)
else:
_writer = new_writer
class ConsoleOutputWriter(object):
def __init__(self, debug=False, quiet=False):
"""
Default output writer that output everything on console.
:param bool debug: print debug messages on standard error
:param bool quiet: don't print info messages
"""
self._debug = debug
self._quiet = quiet
#: Used in check command to hold the check results
self.result_check_list = []
#: Used in status command to hold the status results
self.result_status_list = []
#: The minimal flag. If set the command must output a single list of
#: values.
self.minimal = False
#: The server is active
self.active = True
def _out(self, message, args):
"""
Print a message on standard output
"""
print(_format_message(message, args), file=sys.stdout)
def _err(self, message, args):
"""
Print a message on standard error
"""
print(_format_message(message, args), file=sys.stderr)
def is_quiet(self):
"""
Access the quiet property of the OutputWriter instance
:return bool: if the writer is quiet or not
"""
return self._quiet
def is_debug(self):
"""
Access the debug property of the OutputWriter instance
:return bool: if the writer is in debug mode or not
"""
return self._debug
def debug(self, message, *args):
"""
Emit debug.
"""
if self._debug:
self._err('DEBUG: %s' % message, args)
def info(self, message, *args):
"""
Normal messages are sent to standard output
"""
if not self._quiet:
self._out(message, args)
def warning(self, message, *args):
"""
Warning messages are sent to standard error
"""
self._err('WARNING: %s' % message, args)
def error(self, message, *args):
"""
Error messages are sent to standard error
"""
self._err('ERROR: %s' % message, args)
def exception(self, message, *args):
"""
Warning messages are sent to standard error
"""
self._err('EXCEPTION: %s' % message, args)
def error_occurred(self):
"""
Called immediately before any message method when the originating
call has is_error=True
"""
def close(self):
"""
Close the output channel.
Nothing to do for console.
"""
def result_backup(self, backup_info):
"""
Render the result of a backup.
Nothing to do for console.
"""
# TODO: evaluate to display something useful here
def result_recovery(self, results):
"""
Render the result of a recovery.
"""
if len(results['changes']) > 0:
self.info("")
self.info("IMPORTANT")
self.info("These settings have been modified to prevent "
"data losses")
self.info("")
for assertion in results['changes']:
self.info("%s line %s: %s = %s",
assertion.filename,
assertion.line,
assertion.key,
assertion.value)
if len(results['warnings']) > 0:
self.info("")
self.info("WARNING")
self.info("You are required to review the following options"
" as potentially dangerous")
self.info("")
for assertion in results['warnings']:
self.info("%s line %s: %s = %s",
assertion.filename,
assertion.line,
assertion.key,
assertion.value)
if results['missing_files']:
# At least one file is missing, warn the user
self.info("")
self.info("WARNING")
self.info("The following configuration files have not been "
"saved during backup, hence they have not been "
"restored.")
self.info("You need to manually restore them "
"in order to start the recovered PostgreSQL instance:")
self.info("")
for file_name in results['missing_files']:
self.info(" %s" % file_name)
if results['delete_barman_xlog']:
self.info("")
self.info("After the recovery, please remember to remove the "
"\"barman_xlog\" directory")
self.info("inside the PostgreSQL data directory.")
if results['get_wal']:
self.info("")
self.info("WARNING: 'get-wal' is in the specified "
"'recovery_options'.")
self.info("Before you start up the PostgreSQL server, please "
"review the recovery.conf file")
self.info("inside the target directory. Make sure that "
"'restore_command' can be executed by "
"the PostgreSQL user.")
self.info("")
self.info("Your PostgreSQL server has been successfully "
"prepared for recovery!")
def _record_check(self, server_name, check, status, hint):
"""
Record the check line in result_check_map attribute
This method is for subclass use
:param str server_name: the server is being checked
:param str check: the check name
:param bool status: True if succeeded
:param str,None hint: hint to print if not None
"""
self.result_check_list.append(dict(
server_name=server_name, check=check, status=status, hint=hint))
if not status and self.active:
global error_occurred
error_occurred = True
def init_check(self, server_name, active):
"""
Init the check command
:param str server_name: the server we are start listing
:param boolean active: The server is active
"""
self.info("Server %s:" % server_name)
self.active = active
def result_check(self, server_name, check, status, hint=None):
"""
Record a server result of a server check
and output it as INFO
:param str server_name: the server is being checked
:param str check: the check name
:param bool status: True if succeeded
:param str,None hint: hint to print if not None
"""
self._record_check(server_name, check, status, hint)
if hint:
self.info("\t%s: %s (%s)" %
(check, 'OK' if status else 'FAILED', hint))
else:
self.info("\t%s: %s" %
(check, 'OK' if status else 'FAILED'))
def init_list_backup(self, server_name, minimal=False):
"""
Init the list-backup command
:param str server_name: the server we are start listing
:param bool minimal: if true output only a list of backup id
"""
self.minimal = minimal
def result_list_backup(self, backup_info,
backup_size, wal_size,
retention_status):
"""
Output a single backup in the list-backup command
:param BackupInfo backup_info: backup we are displaying
:param backup_size: size of base backup (with the required WAL files)
:param wal_size: size of WAL files belonging to this backup
(without the required WAL files)
:param retention_status: retention policy status
"""
# If minimal is set only output the backup id
if self.minimal:
self.info(backup_info.backup_id)
return
out_list = [
"%s %s - " % (backup_info.server_name, backup_info.backup_id)]
if backup_info.status == BackupInfo.DONE:
end_time = backup_info.end_time.ctime()
out_list.append('%s - Size: %s - WAL Size: %s' %
(end_time,
pretty_size(backup_size),
pretty_size(wal_size)))
if backup_info.tablespaces:
tablespaces = [("%s:%s" % (tablespace.name,
tablespace.location))
for tablespace in backup_info.tablespaces]
out_list.append(' (tablespaces: %s)' %
', '.join(tablespaces))
if retention_status:
out_list.append(' - %s' % retention_status)
else:
out_list.append(backup_info.status)
self.info(''.join(out_list))
def result_show_backup(self, backup_ext_info):
"""
Output all available information about a backup in show-backup command
The argument has to be the result
of a Server.get_backup_ext_info() call
:param dict backup_ext_info: a dictionary containing
the info to display
"""
data = dict(backup_ext_info)
self.info("Backup %s:", data['backup_id'])
self.info(" Server Name : %s", data['server_name'])
self.info(" Status : %s", data['status'])
if data['status'] == BackupInfo.DONE:
self.info(" PostgreSQL Version : %s", data['version'])
self.info(" PGDATA directory : %s", data['pgdata'])
if data['tablespaces']:
self.info(" Tablespaces:")
for item in data['tablespaces']:
self.info(" %s: %s (oid: %s)",
item.name, item.location, item.oid)
self.info("")
self.info(" Base backup information:")
self.info(" Disk usage : %s (%s with WALs)",
pretty_size(data['size']),
pretty_size(data['size'] + data[
'wal_size']))
if data['deduplicated_size'] is not None and data['size'] > 0:
deduplication_ratio = 1 - (float(data['deduplicated_size']) /
data['size'])
self.info(" Incremental size : %s (-%s)",
pretty_size(data['deduplicated_size']),
'{percent:.2%}'.format(percent=deduplication_ratio)
)
self.info(" Timeline : %s", data['timeline'])
self.info(" Begin WAL : %s",
data['begin_wal'])
self.info(" End WAL : %s", data['end_wal'])
self.info(" WAL number : %s", data['wal_num'])
# Output WAL compression ratio for basebackup WAL files
if data['wal_compression_ratio'] > 0:
self.info(" WAL compression ratio: %s",
'{percent:.2%}'.format(
percent=data['wal_compression_ratio']))
self.info(" Begin time : %s",
data['begin_time'])
self.info(" End time : %s", data['end_time'])
self.info(" Begin Offset : %s",
data['begin_offset'])
self.info(" End Offset : %s",
data['end_offset'])
self.info(" Begin XLOG : %s",
data['begin_xlog'])
self.info(" End XLOG : %s", data['end_xlog'])
self.info("")
self.info(" WAL information:")
self.info(" No of files : %s",
data['wal_until_next_num'])
self.info(" Disk usage : %s",
pretty_size(data['wal_until_next_size']))
# Output WAL rate
if data['wals_per_second'] > 0:
self.info(" WAL rate : %0.2f/hour",
data['wals_per_second'] * 3600)
# Output WAL compression ratio for archived WAL files
if data['wal_until_next_compression_ratio'] > 0:
self.info(
" Compression ratio : %s",
'{percent:.2%}'.format(
percent=data['wal_until_next_compression_ratio']))
self.info(" Last available : %s", data['wal_last'])
if data['children_timelines']:
timelines = data['children_timelines']
self.info(
" Reachable timelines : %s",
", ".join([str(history.tli) for history in timelines]))
self.info("")
self.info(" Catalog information:")
self.info(" Retention Policy : %s",
data['retention_policy_status'] or
'not enforced')
self.info(" Previous Backup : %s",
data.setdefault('previous_backup_id', 'not available') or
'- (this is the oldest base backup)')
self.info(" Next Backup : %s",
data.setdefault('next_backup_id', 'not available') or
'- (this is the latest base backup)')
else:
if data['error']:
self.info(" Error: : %s",
data['error'])
if data['children_timelines']:
self.info("")
self.info(
"WARNING: WAL information is inaccurate due to "
"multiple timelines interacting with this backup")
def init_status(self, server_name):
"""
Init the status command
:param str server_name: the server we are start listing
"""
self.info("Server %s:", server_name)
def result_status(self, server_name, status, description, message):
"""
Record a result line of a server status command
and output it as INFO
:param str server_name: the server is being checked
:param str status: the returned status code
:param str description: the returned status description
:param str,object message: status message. It will be converted to str
"""
message = str(message)
self.result_status_list.append(dict(
server_name=server_name, status=status,
description=description, message=message))
self.info("\t%s: %s", description, message)
def init_replication_status(self, server_name, minimal=False):
"""
Init the 'standby-status' command
:param str server_name: the server we are start listing
:param str minimal: minimal output
"""
self.minimal = minimal
def result_replication_status(self, server_name, target, xlog_location,
standby_info):
"""
Record a result line of a server status command
and output it as INFO
:param str server_name: the replication server
:param str target: all|hot-standby|wal-streamer
:param str xlog_location: server's xlog location
:param StatReplication standby_info: status info of a standby
"""
if target == 'hot-standby':
title = 'hot standby servers'
elif target == 'wal-streamer':
title = 'WAL streamers'
else:
title = 'streaming clients'
if self.minimal:
# Minimal output
if xlog_location:
# xlog location from the master
self.info("%s for master '%s' (xlog @ %s):",
title.capitalize(), server_name, xlog_location)
else:
# We are connected to a standby
self.info("%s for slave '%s':",
title.capitalize(), server_name)
else:
# Full output
self.info("Status of %s for server '%s':",
title, server_name)
# xlog location from the master
if xlog_location:
self.info(" Current xlog location on master: %s",
xlog_location)
if standby_info is not None and not len(standby_info):
self.info(" No %s attached", title)
return
# Minimal output
if self.minimal:
n = 1
for standby in standby_info:
if not standby.replay_location:
# WAL streamer
self.info(" %s. W) %s@%s S:%s W:%s P:%s AN:%s",
n,
standby.usename,
standby.client_addr or 'socket',
standby.sent_location,
standby.write_location,
standby.sync_priority,
standby.application_name)
else:
# Standby
self.info(" %s. %s) %s@%s S:%s F:%s R:%s P:%s AN:%s",
n,
standby.sync_state[0].upper(),
standby.usename,
standby.client_addr or 'socket',
standby.sent_location,
standby.flush_location,
standby.replay_location,
standby.sync_priority,
standby.application_name)
n += 1
else:
n = 1
self.info(" Number of %s: %s",
title, len(standby_info))
for standby in standby_info:
self.info("")
# Calculate differences in bytes
sent_diff = diff_lsn(standby.sent_location,
standby.current_location)
write_diff = diff_lsn(standby.write_location,
standby.current_location)
flush_diff = diff_lsn(standby.flush_location,
standby.current_location)
replay_diff = diff_lsn(standby.replay_location,
standby.current_location)
# Determine the sync stage of the client
sync_stage = None
if not standby.replay_location:
client_type = 'WAL streamer'
max_level = 3
else:
client_type = 'standby'
max_level = 5
# Only standby can replay WAL info
if replay_diff == 0:
sync_stage = '5/5 Hot standby (max)'
elif flush_diff == 0:
sync_stage = '4/5 2-safe' # remote flush
# If not yet done, set the sync stage
if not sync_stage:
if write_diff == 0:
sync_stage = '3/%s Remote write' % max_level
elif sent_diff == 0:
sync_stage = '2/%s WAL Sent (min)' % max_level
else:
sync_stage = '1/%s 1-safe' % max_level
# Synchronous standby
if getattr(standby, 'sync_priority', None) > 0:
self.info(" %s. #%s %s %s",
n,
standby.sync_priority,
standby.sync_state.capitalize(),
client_type)
# Asynchronous standby
else:
self.info(" %s. %s %s",
n,
standby.sync_state.capitalize(),
client_type)
self.info(" Application name: %s",
standby.application_name)
self.info(" Sync stage : %s",
sync_stage)
if getattr(standby, 'client_addr', None):
self.info(" Communication : TCP/IP")
self.info(" IP Address : %s "
"/ Port: %s / Host: %s",
standby.client_addr,
standby.client_port,
standby.client_hostname or '-')
else:
self.info(" Communication : Unix domain socket")
self.info(" User name : %s", standby.usename)
self.info(" Current state : %s (%s)",
standby.state,
standby.sync_state)
if getattr(standby, 'slot_name', None):
self.info(" Replication slot: %s", standby.slot_name)
self.info(" WAL sender PID : %s", standby.pid)
self.info(" Started at : %s", standby.backend_start)
if getattr(standby, 'backend_xmin', None):
self.info(" Standby's xmin : %s",
standby.backend_xmin or '-')
if getattr(standby, 'sent_location', None):
self.info(" Sent location : %s (diff: %s)",
standby.sent_location,
pretty_size(sent_diff))
if getattr(standby, 'write_location', None):
self.info(" Write location : %s (diff: %s)",
standby.write_location,
pretty_size(write_diff))
if getattr(standby, 'flush_location', None):
self.info(" Flush location : %s (diff: %s)",
standby.flush_location,
pretty_size(flush_diff))
if getattr(standby, 'replay_location', None):
self.info(" Replay location : %s (diff: %s)",
standby.replay_location,
pretty_size(replay_diff))
n += 1
def init_list_server(self, server_name, minimal=False):
"""
Init the list-server command
:param str server_name: the server we are start listing
"""
self.minimal = minimal
def result_list_server(self, server_name, description=None):
"""
Output a result line of a list-server command
:param str server_name: the server is being checked
:param str,None description: server description if applicable
"""
if self.minimal or not description:
self.info("%s", server_name)
else:
self.info("%s - %s", server_name, description)
def init_show_server(self, server_name):
"""
Init the show-server command output method
:param str server_name: the server we are displaying
"""
self.info("Server %s:" % server_name)
def result_show_server(self, server_name, server_info):
"""
Output the results of the show-server command
:param str server_name: the server we are displaying
:param dict server_info: a dictionary containing the info to display
"""
for status, message in sorted(server_info.items()):
self.info("\t%s: %s", status, message)
class NagiosOutputWriter(ConsoleOutputWriter):
"""
Nagios output writer.
This writer doesn't output anything to console.
On close it writes a nagios-plugin compatible status
"""
def _out(self, message, args):
"""
Do not print anything on standard output
"""
def _err(self, message, args):
"""
Do not print anything on standard error
"""
def close(self):
"""
Display the result of a check run as expected by Nagios.
Also set the exit code as 2 (CRITICAL) in case of errors
"""
global error_occurred, error_exit_code
# List of all servers that have been checked
servers = []
# List of servers reporting issues
issues = []
for item in self.result_check_list:
# Keep track of all the checked servers
if item['server_name'] not in servers:
servers.append(item['server_name'])
# Keep track of the servers with issues
if not item['status'] and item['server_name'] not in issues:
issues.append(item['server_name'])
# Global error (detected at configuration level)
if len(issues) == 0 and error_occurred:
print("BARMAN CRITICAL - Global configuration errors")
error_exit_code = 2
return
if len(issues) > 0:
fail_summary = []
details = []
for server in issues:
# Join all the issues for a server. Output format is in the
# form:
# " FAILED: , ... "
# All strings will be concatenated into the $SERVICEOUTPUT$
# macro of the Nagios output
server_fail = "%s FAILED: %s" % (
server,
", ".join([
item['check']
for item in self.result_check_list
if item['server_name'] == server and not item['status']
]))
fail_summary.append(server_fail)
# Prepare an array with the detailed output for
# the $LONGSERVICEOUTPUT$ macro of the Nagios output
# line format:
# .: FAILED
# .: FAILED (Hint if present)
# : FAILED
# .....
for issue in self.result_check_list:
if issue['server_name'] == server and not issue['status']:
fail_detail = "%s.%s: FAILED" % (server,
issue['check'])
if issue['hint']:
fail_detail += " (%s)" % issue['hint']
details.append(fail_detail)
# Append the summary of failures to the first line of the output
# using * as delimiter
if len(servers) == 1:
print("BARMAN CRITICAL - server %s has issues * %s" %
(servers[0], " * ".join(fail_summary)))
else:
print("BARMAN CRITICAL - %d server out of %d have issues * "
"%s" % (len(issues), len(servers),
" * ".join(fail_summary)))
# add the detailed list to the output
for issue in details:
print(issue)
error_exit_code = 2
else:
# No issues, all good!
# Display the output message for a single server check
if len(servers) == 1:
print("BARMAN OK - Ready to serve the Espresso backup "
"for %s" %
(servers[0]))
else:
# Display the output message for several servers, using
# '*' as delimiter
print("BARMAN OK - Ready to serve the Espresso backup "
"for %d server(s) * %s" % (
len(servers),
" * ".join([server for server in servers])))
#: This dictionary acts as a registry of available OutputWriters
AVAILABLE_WRITERS = {
'console': ConsoleOutputWriter,
# nagios is not registered as it isn't a general purpose output writer
# 'nagios': NagiosOutputWriter,
}
#: The default OutputWriter
DEFAULT_WRITER = 'console'
#: the current active writer. Initialized according DEFAULT_WRITER on load
_writer = AVAILABLE_WRITERS[DEFAULT_WRITER]()
barman-2.1/barman/fs.py 0000664 0001750 0001750 00000026471 13016764724 015460 0 ustar gcalacoci gcalacoci # Copyright (C) 2013-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
import logging
from barman.command_wrappers import Command
from barman.exceptions import FsOperationFailed
_logger = logging.getLogger(__name__)
def _str(cmd_out):
"""
Make a string from the output of a CommandWrapper execution.
If input is None returns a literal 'None' string
:param cmd_out: String or ByteString to convert
:return str: a string
"""
if hasattr(cmd_out, 'decode') and callable(cmd_out.decode):
return cmd_out.decode('utf-8', 'replace')
else:
return str(cmd_out)
class UnixLocalCommand(object):
"""
This class is a wrapper for local calls for file system operations
"""
def __init__(self, path=None):
# initialize a shell
self.cmd = Command(cmd='sh -c', shell=True, path=path)
def get_last_output(self):
"""
Return the output and the error strings from the last executed command
:rtype: tuple[str,str]
"""
return _str(self.cmd.out), _str(self.cmd.err)
def create_dir_if_not_exists(self, dir_path):
"""
This method check for the existence of a directory.
if exist and is not a directory throws exception.
if is a directory everything is ok and no
mkdir operation is required.
Otherwise creates the directory using mkdir
if the mkdir fails an error is raised
:param dir_path full path for the directory
"""
_logger.debug('Create directory %s if it does not exists' % dir_path)
exists = self.exists(dir_path)
if exists:
is_dir = self.cmd('test -d %s' % dir_path)
if is_dir != 0:
raise FsOperationFailed(
'A file with the same name already exists')
else:
return False
else:
mkdir_ret = self.cmd('mkdir %s' % dir_path)
if mkdir_ret == 0:
return True
else:
raise FsOperationFailed('mkdir execution failed')
def delete_if_exists(self, path):
"""
This method check for the existence of a path.
If it exists, then is removed using a rm -fr command,
and returns True.
If the command fails an exception is raised.
If the path does not exists returns False
:param path the full path for the directory
"""
_logger.debug('Delete path %s if exists' % path)
exists = self.exists(path, False)
if exists:
rm_ret = self.cmd('rm -fr %s' % path)
if rm_ret == 0:
return True
else:
raise FsOperationFailed('rm execution failed')
else:
return False
def check_directory_exists(self, dir_path):
"""
Check for the existence of a directory in path.
if the directory exists returns true.
if the directory does not exists returns false.
if exists a file and is not a directory raises an exception
:param dir_path full path for the directory
"""
_logger.debug('Check if directory %s exists' % dir_path)
exists = self.exists(dir_path)
if exists:
is_dir = self.cmd('test -d %s' % dir_path)
if is_dir != 0:
raise FsOperationFailed(
'A file with the same name exists, but is not a directory')
else:
return True
else:
return False
def check_write_permission(self, dir_path):
"""
check write permission for barman on a given path.
Creates a hidden file using touch, then remove the file.
returns true if the file is written and removed without problems
raise exception if the creation fails.
raise exception if the removal fails.
:param dir_path full dir_path for the directory to check
"""
_logger.debug('Check if directory %s is writable' % dir_path)
exists = self.exists(dir_path)
if exists:
is_dir = self.cmd('test -d %s' % dir_path)
if is_dir == 0:
can_write = self.cmd('touch %s/.barman_write_check' % dir_path)
if can_write == 0:
can_remove = self.cmd(
'rm %s/.barman_write_check' % dir_path)
if can_remove == 0:
return True
else:
raise FsOperationFailed('Unable to remove file')
else:
raise FsOperationFailed(
'Unable to create write check file')
else:
raise FsOperationFailed('%s is not a directory' % dir_path)
else:
raise FsOperationFailed('%s does not exists' % dir_path)
def create_symbolic_link(self, src, dst):
"""
Create a symlink pointing to src named dst.
Check src exists, if so, checks that destination
does not exists. if src is an invalid folder, raises an exception.
if dst already exists, raises an exception. if ln -s command fails
raises an exception
:param src full path to the source of the symlink
:param dst full path for the destination of the symlink
"""
_logger.debug('Create symbolic link %s -> %s' % (src, dst))
exists = self.exists(src)
if exists:
exists_dst = self.exists(dst)
if not exists_dst:
link = self.cmd('ln -s %s %s' % (src, dst))
if link == 0:
return True
else:
raise FsOperationFailed('ln command failed')
else:
raise FsOperationFailed('ln destination already exists')
else:
raise FsOperationFailed('ln source does not exists')
def get_system_info(self):
"""
Gather important system information for 'barman diagnose' command
"""
result = {}
# self.cmd.out can be None. The str() call will ensure it will be
# translated to a literal 'None'
release = ''
if self.cmd("lsb_release -a") == 0:
release = _str(self.cmd.out).rstrip()
elif self.exists('/etc/lsb-release'):
self.cmd('cat /etc/lsb-release ')
release = "Ubuntu Linux %s" % _str(self.cmd.out).rstrip()
elif self.exists('/etc/debian_version'):
self.cmd('cat /etc/debian_version')
release = "Debian GNU/Linux %s" % _str(self.cmd.out).rstrip()
elif self.exists('/etc/redhat-release'):
self.cmd('cat /etc/redhat-release')
release = "RedHat Linux %s" % _str(self.cmd.out).rstrip()
elif self.cmd('sw_vers') == 0:
release = _str(self.cmd.out).rstrip()
result['release'] = release
self.cmd('uname -a')
result['kernel_ver'] = _str(self.cmd.out).rstrip()
self.cmd('python --version 2>&1')
result['python_ver'] = _str(self.cmd.out).rstrip()
self.cmd('rsync --version 2>&1')
result['rsync_ver'] = _str(self.cmd.out).splitlines(True)[0].rstrip()
self.cmd('ssh -V 2>&1')
result['ssh_ver'] = _str(self.cmd.out).rstrip()
return result
def get_file_content(self, path):
"""
Retrieve the content of a file
If the file doesn't exist or isn't readable, it raises an exception.
:param str path: full path to the file to read
"""
_logger.debug('Reading content of file %s' % path)
result = self.exists(path)
if not result:
raise FsOperationFailed('The %s file does not exist' % path)
result = self.cmd("test -r '%s'" % path)
if result != 0:
raise FsOperationFailed('The %s file is not readable' % path)
result = self.cmd("cat '%s'" % path)
if result != 0:
raise FsOperationFailed('Failed to execute "cat \'%s\'"' % path)
return self.cmd.out
def exists(self, path, dereference=True):
"""
Check for the existence of a path.
:param str path: full path to check
:param bool dereference: whether dereference symlinks, defaults
to True
:return bool: if the file exists or not.
"""
_logger.debug('check for existence of: %s' % path)
cmd_str = "test -e '%s'" % path
if not dereference:
cmd_str += " -o -L '%s'" % path
result = self.cmd(cmd_str)
return result == 0
def ping(self):
"""
'Ping' the server executing the `true` command.
:return int: the true cmd result
"""
_logger.debug('execute the true command')
result = self.cmd("true")
return result
def list_dir_content(self, dir_path, options=''):
"""
List the contents of a given directory.
:param str dir_path: the path where we want the ls to be executed
:param str options: a string containing the options for the ls command
:return str: the ls cmd output
"""
_logger.debug('list the content of a directory')
ls_command = "ls %s '%s'" % (options, dir_path)
self.cmd(ls_command)
return self.cmd.out
class UnixRemoteCommand(UnixLocalCommand):
"""
This class is a wrapper for remote calls for file system operations
"""
# noinspection PyMissingConstructor
def __init__(self, ssh_command, ssh_options=None, path=None):
"""
Uses the same commands as the UnixLocalCommand
but the constructor is overridden and a remote shell is
initialized using the ssh_command provided by the user
:param str ssh_command: the ssh command provided by the user
:param list[str] ssh_options: the options to be passed to SSH
:param str path: the path to be used if provided, otherwise
the PATH environment variable will be used
"""
# Ensure that ssh_option is iterable
if ssh_options is None:
ssh_options = []
if ssh_command is None:
raise FsOperationFailed('No ssh command provided')
self.cmd = Command(ssh_command,
ssh_options,
path=path,
shell=True)
try:
ret = self.cmd("true")
except OSError:
raise FsOperationFailed("Unable to execute %s" % ssh_command)
if ret != 0:
raise FsOperationFailed(
"Connection failed using '%s %s' return code %s" % (
ssh_command,
' '.join(ssh_options),
ret))
barman-2.1/barman/xlog.py 0000664 0001750 0001750 00000027171 13020304015 015771 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module contains functions to retrieve information about xlog
files
"""
import collections
import os
import re
from tempfile import NamedTemporaryFile
from barman.exceptions import BadHistoryFileContents, BadXlogSegmentName
# xlog file segment name parser (regular expression)
_xlog_re = re.compile(r'''
^
([\dA-Fa-f]{8}) # everything has a timeline
(?:
([\dA-Fa-f]{8})([\dA-Fa-f]{8}) # segment name, if a wal file
(?: # and optional
\.[\dA-Fa-f]{8}\.backup # offset, if a backup label
|
\.partial # partial, if a partial file
)?
|
\.history # or only .history, if a history file
)
$
''', re.VERBOSE)
# xlog location parser for concurrent backup (regular expression)
_location_re = re.compile(r'^([\dA-F]+)/([\dA-F]+)$')
# Taken from xlog_internal.h from PostgreSQL sources
XLOG_SEG_SIZE = 1 << 24
XLOG_SEG_PER_FILE = 0xffffffff // XLOG_SEG_SIZE
XLOG_FILE_SIZE = XLOG_SEG_SIZE * XLOG_SEG_PER_FILE
#: This namedtuple is a container for the information
#: contained inside history files
HistoryFileData = collections.namedtuple(
'HistoryFileData',
'tli parent_tli switchpoint reason')
def is_any_xlog_file(path):
"""
Return True if the xlog is either a WAL segment, a .backup file
or a .history file, False otherwise.
It supports either a full file path or a simple file name.
:param str path: the file name to test
:rtype: bool
"""
match = _xlog_re.match(os.path.basename(path))
if match:
return True
return False
def is_history_file(path):
"""
Return True if the xlog is a .history file, False otherwise
It supports either a full file path or a simple file name.
:param str path: the file name to test
:rtype: bool
"""
match = _xlog_re.search(os.path.basename(path))
if match and match.group(0).endswith('.history'):
return True
return False
def is_backup_file(path):
"""
Return True if the xlog is a .backup file, False otherwise
It supports either a full file path or a simple file name.
:param str path: the file name to test
:rtype: bool
"""
match = _xlog_re.search(os.path.basename(path))
if match and match.group(0).endswith('.backup'):
return True
return False
def is_partial_file(path):
"""
Return True if the xlog is a .partial file, False otherwise
It supports either a full file path or a simple file name.
:param str path: the file name to test
:rtype: bool
"""
match = _xlog_re.search(os.path.basename(path))
if match and match.group(0).endswith('.partial'):
return True
return False
def is_wal_file(path):
"""
Return True if the xlog is a regular xlog file, False otherwise
It supports either a full file path or a simple file name.
:param str path: the file name to test
:rtype: bool
"""
match = _xlog_re.search(os.path.basename(path))
if (match and
not match.group(0).endswith('.backup') and
not match.group(0).endswith('.history') and
not match.group(0).endswith('.partial')):
return True
return False
def decode_segment_name(path):
"""
Retrieve the timeline, log ID and segment ID
from the name of a xlog segment
It can handle either a full file path or a simple file name.
:param str path: the file name to decode
:rtype: list[int]
"""
name = os.path.basename(path)
match = _xlog_re.match(name)
if not match:
raise BadXlogSegmentName(name)
return [int(x, 16) if x else None for x in match.groups()]
def encode_segment_name(tli, log, seg):
"""
Build the xlog segment name based on timeline, log ID and segment ID
:param int tli: timeline number
:param int log: log number
:param int seg: segment number
:return str: segment file name
"""
return "%08X%08X%08X" % (tli, log, seg)
def encode_history_file_name(tli):
"""
Build the history file name based on timeline
:return str: history file name
"""
return "%08X.history" % (tli,)
def generate_segment_names(begin, end=None, version=None):
"""
Generate a sequence of XLOG segments starting from ``begin``
If an ``end`` segment is provided the sequence will terminate after
returning it, otherwise the sequence will never terminate.
:param str begin: begin segment name
:param str|None end: optional end segment name
:param int|None version: optional postgres version as an integer
(e.g. 90301 for 9.3.1)
:rtype: collections.Iterable[str]
:raise: BadXlogSegmentName
"""
begin_tli, begin_log, begin_seg = decode_segment_name(begin)
end_tli, end_log, end_seg = None, None, None
if end:
end_tli, end_log, end_seg = decode_segment_name(end)
# this method doesn't support timeline changes
assert begin_tli == end_tli, (
"Begin segment (%s) and end segment (%s) "
"must have the same timeline part" % (begin, end))
# If version is less than 9.3 the last segment must be skipped
skip_last_segment = version is not None and version < 90300
# Start from the first xlog and generate the segments sequentially
# If ``end`` has been provided, the while condition ensure the termination
# otherwise this generator will never stop
cur_log, cur_seg = begin_log, begin_seg
while end is None or \
cur_log < end_log or \
(cur_log == end_log and cur_seg <= end_seg):
yield encode_segment_name(begin_tli, cur_log, cur_seg)
cur_seg += 1
if cur_seg > XLOG_SEG_PER_FILE or (
skip_last_segment and cur_seg == XLOG_SEG_PER_FILE):
cur_seg = 0
cur_log += 1
def hash_dir(path):
"""
Get the directory where the xlog segment will be stored
It can handle either a full file path or a simple file name.
:param str|unicode path: xlog file name
:return str: directory name
"""
tli, log, _ = decode_segment_name(path)
# tli is always not None
if log is not None:
return "%08X%08X" % (tli, log)
else:
return ''
def parse_lsn(lsn_string):
"""
Transform a string XLOG location, formatted as %X/%X, in the corresponding
numeric representation
:param str lsn_string: the string XLOG location, i.e. '2/82000168'
:rtype: int
"""
lsn_list = lsn_string.split('/')
if len(lsn_list) != 2:
raise ValueError('Invalid LSN: %s', lsn_string)
return (int(lsn_list[0], 16) << 32) + int(lsn_list[1], 16)
def diff_lsn(lsn_string1, lsn_string2):
"""
Calculate the difference in bytes between two string XLOG location,
formatted as %X/%X
Tis function is a Python implementation of
the ``pg_xlog_location_diff(str, str)`` PostgreSQL function.
:param str lsn_string1: the string XLOG location, i.e. '2/82000168'
:param str lsn_string2: the string XLOG location, i.e. '2/82000168'
:rtype: int
"""
# If one the input is None returns None
if lsn_string1 is None or lsn_string2 is None:
return None
return parse_lsn(lsn_string1) - parse_lsn(lsn_string2)
def format_lsn(lsn):
"""
Transform a numeric XLOG location, in the corresponding %X/%X string
representation
:param int lsn: numeric XLOG location
:rtype: str
"""
return "%X/%X" % (lsn >> 32, lsn & 0xFFFFFFFF)
def location_to_xlogfile_name_offset(location, timeline):
"""
Convert transaction log location string to file_name and file_offset
This is a reimplementation of pg_xlogfile_name_offset PostgreSQL function
This method returns a dictionary containing the following data:
* file_name
* file_offset
:param str location: XLOG location
:param int timeline: timeline
:rtype: dict
"""
lsn = parse_lsn(location)
log = lsn >> 32
seg = (lsn & XLOG_FILE_SIZE) >> 24
offset = lsn & 0xFFFFFF
return {
'file_name': encode_segment_name(timeline, log, seg),
'file_offset': offset,
}
def location_from_xlogfile_name_offset(file_name, file_offset):
"""
Convert file_name and file_offset to a transaction log location.
This is the inverted function of PostgreSQL's pg_xlogfile_name_offset
function.
:param str file_name: a WAL file name
:param int file_offset: a numeric offset
:rtype: str
"""
decoded_segment = decode_segment_name(file_name)
location = ((decoded_segment[1] << 32) +
(decoded_segment[2] << 24) +
file_offset)
return format_lsn(location)
def decode_history_file(wal_info, comp_manager):
"""
Read an history file and parse its contents.
Each line in the file represents a timeline switch, each field is
separated by tab, empty lines are ignored and lines starting with '#'
are comments.
Each line is composed by three fields: parentTLI, switchpoint and reason.
"parentTLI" is the ID of the parent timeline.
"switchpoint" is the WAL position where the switch happened
"reason" is an human-readable explanation of why the timeline was changed
The method requires a CompressionManager object to handle the eventual
compression of the history file.
:param barman.infofile.WalFileInfo wal_info: history file obj
:param comp_manager: compression manager used in case
of history file compression
:return List[HistoryFileData]: information from the history file
"""
path = wal_info.orig_filename
# Decompress the file if needed
if wal_info.compression:
# Use a NamedTemporaryFile to avoid explicit cleanup
uncompressed_file = NamedTemporaryFile(
dir=os.path.dirname(path),
prefix='.%s.' % wal_info.name,
suffix='.uncompressed')
path = uncompressed_file.name
comp_manager.get_compressor(wal_info.compression).decompress(
wal_info.orig_filename, path)
# Extract the timeline from history file name
tli, _, _ = decode_segment_name(wal_info.name)
lines = []
with open(path) as fp:
for line in fp:
line = line.strip()
# Skip comments and empty lines
if line.startswith("#"):
continue
# Skip comments and empty lines
if len(line) == 0:
continue
# Use tab as separator
contents = line.split('\t')
if len(contents) != 3:
# Invalid content of the line
raise BadHistoryFileContents(path)
history = HistoryFileData(
tli=tli,
parent_tli=int(contents[0]),
switchpoint=parse_lsn(contents[1]),
reason=contents[2])
lines.append(history)
# Empty history file or containing invalid content
if len(lines) == 0:
raise BadHistoryFileContents(path)
else:
return lines
barman-2.1/barman/compression.py 0000664 0001750 0001750 00000024607 13015546105 017376 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module is responsible to manage the compression features of Barman
"""
import bz2
import gzip
import logging
import shutil
from abc import ABCMeta, abstractmethod
from contextlib import closing
from barman.command_wrappers import Command
from barman.exceptions import (CommandFailedException,
CompressionIncompatibility)
from barman.utils import with_metaclass
_logger = logging.getLogger(__name__)
class CompressionManager(object):
def __init__(self, config, path):
"""
Compression manager
"""
self.config = config
self.path = path
def check(self, compression=None):
"""
This method returns True if the compression specified in the
configuration file is present in the register, otherwise False
"""
if not compression:
compression = self.config.compression
if compression not in compression_registry:
return False
return True
def get_compressor(self, compression=None):
"""
Returns a new compressor instance
:param str compression: Compression name
"""
if not compression:
compression = self.config.compression
# Check if the requested compression mechanism is allowed
if self.check(compression):
return compression_registry[compression](
config=self.config, compression=compression, path=self.path)
else:
return None
def identify_compression(filename):
"""
Try to guess the compression algorithm of a file
:param filename: the pat of the file to identify
:rtype: str
"""
# TODO: manage multiple decompression methods for the same
# compression algorithm (e.g. what to do when gzip is detected?
# should we use gzip or pigz?)
with open(filename, 'rb') as f:
file_start = f.read(MAGIC_MAX_LENGTH)
for file_type, cls in sorted(compression_registry.items()):
if cls.validate(file_start):
return file_type
return None
class Compressor(with_metaclass(ABCMeta, object)):
"""
Base class for all the compressors
"""
MAGIC = None
def __init__(self, config, compression, path=None):
self.config = config
self.compression = compression
self.path = path
@classmethod
def validate(cls, file_start):
"""
Guess if the first bytes of a file are compatible with the compression
implemented by this class
:param file_start: a binary string representing the first few
bytes of a file
:rtype: bool
"""
return cls.MAGIC and file_start.startswith(cls.MAGIC)
@abstractmethod
def compress(self, src, dst):
"""
Abstract Method for compression method
:param str src: source file path
:param str dst: destination file path
"""
@abstractmethod
def decompress(self, src, dst):
"""
Abstract method for decompression method
:param str src: source file path
:param str dst: destination file path
"""
class CommandCompressor(Compressor):
"""
Base class for compressors built on external commands
"""
def __init__(self, config, compression, path=None):
super(CommandCompressor, self).__init__(
config, compression, path)
self._compress = None
self._decompress = None
def compress(self, src, dst):
"""
Compress using the specific command defined in the sublcass
:param src: source file to compress
:param dst: destination of the decompression
"""
return self._compress(src, dst)
def decompress(self, src, dst):
"""
Decompress using the specific command defined in the sublcass
:param src: source file to decompress
:param dst: destination of the decompression
"""
return self._decompress(src, dst)
def _build_command(self, pipe_command):
"""
Build the command string and create the actual Command object
:param pipe_command: the command used to compress/decompress
:rtype: Command
"""
command = 'command(){ '
command += pipe_command
command += ' > "$2" < "$1"'
command += ';}; command'
return Command(command, shell=True, check=True, path=self.path)
class InternalCompressor(Compressor):
"""
Base class for compressors built on python libraries
"""
def compress(self, src, dst):
"""
Compress using the object defined in the sublcass
:param src: source file to compress
:param dst: destination of the decompression
"""
try:
with open(src, 'rb') as istream:
with closing(self._compressor(dst)) as ostream:
shutil.copyfileobj(istream, ostream)
except Exception as e:
# you won't get more information from the compressors anyway
raise CommandFailedException(dict(ret=None, err=str(e), out=None))
return 0
def decompress(self, src, dst):
"""
Decompress using the object defined in the sublcass
:param src: source file to decompress
:param dst: destination of the decompression
"""
try:
with closing(self._decompressor(src)) as istream:
with open(dst, 'wb') as ostream:
shutil.copyfileobj(istream, ostream)
except Exception as e:
# you won't get more information from the compressors anyway
raise CommandFailedException(dict(ret=None, err=str(e), out=None))
return 0
@abstractmethod
def _decompressor(self, src):
"""
Abstract decompressor factory method
:param src: source file path
:return: a file-like readable decompressor object
"""
@abstractmethod
def _compressor(self, dst):
"""
Abstract compressor factory method
:param dst: destination file path
:return: a file-like writable compressor object
"""
class GZipCompressor(CommandCompressor):
"""
Predefined compressor with GZip
"""
MAGIC = b'\x1f\x8b\x08'
def __init__(self, config, compression, path=None):
super(GZipCompressor, self).__init__(
config, compression, path)
self._compress = self._build_command('gzip -c')
self._decompress = self._build_command('gzip -c -d')
class PyGZipCompressor(InternalCompressor):
"""
Predefined compressor that uses GZip Python libraries
"""
MAGIC = b'\x1f\x8b\x08'
def __init__(self, config, compression, path=None):
super(PyGZipCompressor, self).__init__(
config, compression, path)
# Default compression level used in system gzip utility
self._level = -1 # Z_DEFAULT_COMPRESSION constant of zlib
def _compressor(self, name):
return gzip.GzipFile(name, mode='wb', compresslevel=self._level)
def _decompressor(self, name):
return gzip.GzipFile(name, mode='rb')
class PigzCompressor(CommandCompressor):
"""
Predefined compressor with Pigz
Note that pigz on-disk is the same as gzip, so the MAGIC value of this
class is the same
"""
MAGIC = b'\x1f\x8b\x08'
def __init__(self, config, compression, path=None):
super(PigzCompressor, self).__init__(
config, compression, path)
self._compress = self._build_command('pigz -c')
self._decompress = self._build_command('pigz -c -d')
class BZip2Compressor(CommandCompressor):
"""
Predefined compressor with BZip2
"""
MAGIC = b'\x42\x5a\x68'
def __init__(self, config, compression, path=None):
super(BZip2Compressor, self).__init__(
config, compression, path)
self._compress = self._build_command('bzip2 -c')
self._decompress = self._build_command('bzip2 -c -d')
class PyBZip2Compressor(InternalCompressor):
"""
Predefined compressor with BZip2 Python libraries
"""
MAGIC = b'\x42\x5a\x68'
def __init__(self, config, compression, path=None):
super(PyBZip2Compressor, self).__init__(
config, compression, path)
# Default compression level used in system gzip utility
self._level = 9
def _compressor(self, name):
return bz2.BZ2File(name, mode='wb', compresslevel=self._level)
def _decompressor(self, name):
return bz2.BZ2File(name, mode='rb')
class CustomCompressor(CommandCompressor):
"""
Custom compressor
"""
def __init__(self, config, compression, path=None):
if not config.custom_compression_filter:
raise CompressionIncompatibility("custom_compression_filter")
if not config.custom_decompression_filter:
raise CompressionIncompatibility("custom_decompression_filter")
super(CustomCompressor, self).__init__(
config, compression, path)
self._compress = self._build_command(
config.custom_compression_filter)
self._decompress = self._build_command(
config.custom_decompression_filter)
# a dictionary mapping all supported compression schema
# to the class implementing it
# WARNING: items in this dictionary are extracted using alphabetical order
# It's important that gzip and bzip2 are positioned before their variants
compression_registry = {
'gzip': GZipCompressor,
'pigz': PigzCompressor,
'bzip2': BZip2Compressor,
'pygzip': PyGZipCompressor,
'pybzip2': PyBZip2Compressor,
'custom': CustomCompressor,
}
#: The longest string needed to identify a compression schema
MAGIC_MAX_LENGTH = max(len(x.MAGIC or '')
for x in compression_registry.values())
barman-2.1/barman/cli.py 0000664 0001750 0001750 00000107305 13026732452 015605 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module implements the interface with the command line and the logger.
"""
import logging
import os
import sys
from argparse import SUPPRESS, ArgumentTypeError
from contextlib import closing
from argh import ArghParser, arg, expects_obj, named
import barman.config
import barman.diagnose
from barman import output
from barman.exceptions import BadXlogSegmentName
from barman.infofile import BackupInfo
from barman.server import Server
from barman.utils import configure_logging, drop_privileges, parse_log_level
_logger = logging.getLogger(__name__)
def check_non_negative(value):
"""
Check for a positive integer option
:param value: str containing the value to check
"""
if value is None:
return None
try:
int_value = int(value)
except Exception:
raise ArgumentTypeError("'%s' is not a valid non negative integer" %
value)
if int_value < 0:
raise ArgumentTypeError("'%s' is not a valid non negative integer" %
value)
return int_value
def check_positive(value):
"""
Check for a positive integer option
:param value: str containing the value to check
"""
if value is None:
return None
try:
int_value = int(value)
except Exception:
raise ArgumentTypeError("'%s' is not a valid positive integer" %
value)
if int_value < 1:
raise ArgumentTypeError("'%s' is not a valid positive integer" %
value)
return int_value
@named('list-server')
@arg('--minimal', help='machine readable output')
def list_server(minimal=False):
"""
List available servers, with useful information
"""
# Get every server, both inactive and temporarily disabled
servers = get_server_list()
for name in sorted(servers):
server = servers[name]
# Exception: manage_server_command is not invoked here
# Normally you would call manage_server_command to check if the
# server is None and to report inactive and disabled servers, but here
# we want all servers and the server cannot be None
output.init('list_server', name, minimal=minimal)
description = server.config.description
# If the server has been manually disabled
if not server.config.active:
description += " (inactive)"
# If server has configuration errors
elif server.config.disabled:
description += " (WARNING: disabled)"
output.result('list_server', name, description)
output.close_and_exit()
def cron():
"""
Run maintenance tasks (global command)
"""
# Skip inactive and temporarily disabled servers
servers = get_server_list(skip_inactive=True, skip_disabled=True)
for name in sorted(servers):
server = servers[name]
# Exception: manage_server_command is not invoked here
# Normally you would call manage_server_command to check if the
# server is None and to report inactive and disabled servers,
# but here we have only active and well configured servers.
server.cron()
output.close_and_exit()
# noinspection PyUnusedLocal
def server_completer(prefix, parsed_args, **kwargs):
global_config(parsed_args)
for conf in barman.__config__.servers():
if conf.name.startswith(prefix):
yield conf.name
# noinspection PyUnusedLocal
def server_completer_all(prefix, parsed_args, **kwargs):
global_config(parsed_args)
current_list = getattr(parsed_args, 'server_name', None) or ()
for conf in barman.__config__.servers():
if conf.name.startswith(prefix) and conf.name not in current_list:
yield conf.name
if len(current_list) == 0 and 'all'.startswith(prefix):
yield 'all'
# noinspection PyUnusedLocal
def backup_completer(prefix, parsed_args, **kwargs):
global_config(parsed_args)
server = get_server(parsed_args)
backups = server.get_available_backups()
for backup_id in sorted(backups, reverse=True):
if backup_id.startswith(prefix):
yield backup_id
for special_id in ('latest', 'last', 'oldest', 'first'):
if len(backups) > 0 and special_id.startswith(prefix):
yield special_id
@arg('server_name', nargs='+',
completer=server_completer_all,
help="specifies the server names for the backup command "
"('all' will show all available servers)")
@arg('--immediate-checkpoint',
help='forces the initial checkpoint to be done as quickly as possible',
dest='immediate_checkpoint',
action='store_true',
default=SUPPRESS)
@arg('--no-immediate-checkpoint',
help='forces the initial checkpoint to be spreaded',
dest='immediate_checkpoint',
action='store_false',
default=SUPPRESS)
@arg('--reuse-backup', nargs='?',
choices=barman.config.REUSE_BACKUP_VALUES,
default=None, const='link',
help='use the previous backup to improve transfer-rate. '
'If no argument is given "link" is assumed')
@arg('--retry-times',
help='Number of retries after an error if base backup copy fails.',
type=check_non_negative)
@arg('--retry-sleep',
help='Wait time after a failed base backup copy, before retrying.',
type=check_non_negative)
@arg('--no-retry', help='Disable base backup copy retry logic.',
dest='retry_times', action='store_const', const=0)
@expects_obj
def backup(args):
"""
Perform a full backup for the given server (supports 'all')
"""
servers = get_server_list(args, skip_inactive=True)
for name in sorted(servers):
server = servers[name]
# Skip the server (apply general rule)
if not manage_server_command(server, name):
continue
if args.reuse_backup is not None:
server.config.reuse_backup = args.reuse_backup
if args.retry_sleep is not None:
server.config.basebackup_retry_sleep = args.retry_sleep
if args.retry_times is not None:
server.config.basebackup_retry_times = args.retry_times
if hasattr(args, 'immediate_checkpoint'):
server.config.immediate_checkpoint = args.immediate_checkpoint
with closing(server):
server.backup()
output.close_and_exit()
@named('list-backup')
@arg('server_name', nargs='+',
completer=server_completer_all,
help="specifies the server name for the command "
"('all' will show all available servers)")
@arg('--minimal', help='machine readable output', action='store_true')
@expects_obj
def list_backup(args):
"""
List available backups for the given server (supports 'all')
"""
servers = get_server_list(args, skip_inactive=True)
for name in sorted(servers):
server = servers[name]
# Skip the server (apply general rule)
if not manage_server_command(server, name):
continue
output.init('list_backup', name, minimal=args.minimal)
with closing(server):
server.list_backups()
output.close_and_exit()
@arg('server_name', nargs='+',
completer=server_completer_all,
help='specifies the server name for the command')
@expects_obj
def status(args):
"""
Shows live information and status of the PostgreSQL server
"""
servers = get_server_list(args, skip_inactive=True)
for name in sorted(servers):
server = servers[name]
# Skip the server (apply general rule)
if not manage_server_command(server, name):
continue
output.init('status', name)
with closing(server):
server.status()
output.close_and_exit()
@named('replication-status')
@arg('server_name', nargs='+',
completer=server_completer_all,
help='specifies the server name for the command')
@arg('--minimal', help='machine readable output', action='store_true')
@arg('--target', choices=('all', 'hot-standby', 'wal-streamer'),
default='all',
help='''
Possible values are: 'hot-standby' (only hot standby servers),
'wal-streamer' (only WAL streaming clients, such as pg_receivexlog),
'all' (any of them). Defaults to %(default)s''')
@expects_obj
def replication_status(args):
"""
Shows live information and status of any streaming client
"""
servers = get_server_list(args, skip_inactive=True)
for name in sorted(servers):
server = servers[name]
# Skip the server (apply general rule)
if not manage_server_command(server, name):
continue
with closing(server):
output.init('replication_status',
name,
minimal=args.minimal)
server.replication_status(args.target)
output.close_and_exit()
@arg('server_name', nargs='+',
completer=server_completer_all,
help='specifies the server name for the command')
@expects_obj
def rebuild_xlogdb(args):
"""
Rebuild the WAL file database guessing it from the disk content.
"""
servers = get_server_list(args, skip_inactive=True)
for name in sorted(servers):
server = servers[name]
# Skip the server (apply general rule)
if not manage_server_command(server, name):
continue
with closing(server):
server.rebuild_xlogdb()
output.close_and_exit()
@arg('server_name',
completer=server_completer,
help='specifies the server name for the command')
@arg('--target-tli', help='target timeline', type=check_positive)
@arg('--target-time',
help='target time. You can use any valid unambiguous representation. '
'e.g: "YYYY-MM-DD HH:MM:SS.mmm"')
@arg('--target-xid', help='target transaction ID')
@arg('--target-name',
help='target name created previously with '
'pg_create_restore_point() function call')
@arg('--exclusive',
help='set target xid to be non inclusive', action="store_true")
@arg('--tablespace',
help='tablespace relocation rule',
metavar='NAME:LOCATION', action='append')
@arg('--remote-ssh-command',
metavar='SSH_COMMAND',
help='This options activates remote recovery, by specifying the secure '
'shell command to be launched on a remote host. It is '
'the equivalent of the "ssh_command" server option in '
'the configuration file for remote recovery. '
'Example: "ssh postgres@db2"')
@arg('backup_id',
completer=backup_completer,
help='specifies the backup ID to recover')
@arg('destination_directory',
help='the directory where the new server is created')
@arg('--retry-times',
help='Number of retries after an error if base backup copy fails.',
type=check_non_negative)
@arg('--retry-sleep',
help='Wait time after a failed base backup copy, before retrying.',
type=check_non_negative)
@arg('--no-retry', help='Disable base backup copy retry logic.',
dest='retry_times', action='store_const', const=0)
@expects_obj
def recover(args):
"""
Recover a server at a given time or xid
"""
server = get_server(args)
# Retrieves the backup
backup_id = parse_backup_id(server, args)
if backup_id.status != BackupInfo.DONE:
output.error(
"Cannot recover from backup '%s' of server '%s': "
"backup status is not DONE",
args.backup_id, server.config.name)
output.close_and_exit()
# decode the tablespace relocation rules
tablespaces = {}
if args.tablespace:
for rule in args.tablespace:
try:
tablespaces.update([rule.split(':', 1)])
except ValueError:
output.error(
"Invalid tablespace relocation rule '%s'\n"
"HINT: The valid syntax for a relocation rule is "
"NAME:LOCATION", rule)
output.close_and_exit()
# validate the rules against the tablespace list
valid_tablespaces = []
if backup_id.tablespaces:
valid_tablespaces = [tablespace_data.name for tablespace_data in
backup_id.tablespaces]
for item in tablespaces:
if item not in valid_tablespaces:
output.error("Invalid tablespace name '%s'\n"
"HINT: Please use any of the following "
"tablespaces: %s",
item, ', '.join(valid_tablespaces))
output.close_and_exit()
# explicitly disallow the rsync remote syntax (common mistake)
if ':' in args.destination_directory:
output.error(
"The destination directory parameter "
"cannot contain the ':' character\n"
"HINT: If you want to do a remote recovery you have to use "
"the --remote-ssh-command option")
output.close_and_exit()
if args.retry_sleep is not None:
server.config.basebackup_retry_sleep = args.retry_sleep
if args.retry_times is not None:
server.config.basebackup_retry_times = args.retry_times
with closing(server):
server.recover(backup_id,
args.destination_directory,
tablespaces=tablespaces,
target_tli=args.target_tli,
target_time=args.target_time,
target_xid=args.target_xid,
target_name=args.target_name,
exclusive=args.exclusive,
remote_command=args.remote_ssh_command)
output.close_and_exit()
@named('show-server')
@arg('server_name', nargs='+',
completer=server_completer_all,
help="specifies the server names to show "
"('all' will show all available servers)")
@expects_obj
def show_server(args):
"""
Show all configuration parameters for the specified servers
"""
servers = get_server_list(args)
for name in sorted(servers):
server = servers[name]
# Skip the server (apply general rule)
if not manage_server_command(
server, name, skip_inactive=False,
skip_disabled=False, disabled_is_error=False):
continue
# If the server has been manually disabled
if not server.config.active:
name += " (inactive)"
# If server has configuration errors
elif server.config.disabled:
name += " (WARNING: disabled)"
output.init('show_server', name)
with closing(server):
server.show()
output.close_and_exit()
@named('switch-xlog')
@arg('server_name', nargs='+',
completer=server_completer_all,
help="specifies the server name target of the switch-xlog command")
@arg('--force',
help='forces the switch of a xlog by executing a checkpoint before',
dest='force',
action='store_true',
default=False)
@arg('--archive',
help='wait for one xlog file to be archived',
dest='archive',
action='store_true',
default=False)
@arg('--archive-timeout',
help='the time, in seconds, the archiver will wait for a new xlog file '
'to be archived before timing out',
metavar='TIMEOUT',
default='30',
type=check_non_negative)
@expects_obj
def switch_xlog(args):
"""
Execute the switch-xlog command on the target server
"""
servers = get_server_list(args, skip_inactive=True)
for name in sorted(servers):
server = servers[name]
# Skip the server (apply general rule)
if not manage_server_command(server, name):
continue
with closing(server):
server.switch_xlog(args.force, args.archive, args.archive_timeout)
output.close_and_exit()
@arg('server_name', nargs='+',
completer=server_completer_all,
help="specifies the server names to check "
"('all' will check all available servers)")
@arg('--nagios', help='Nagios plugin compatible output', action='store_true')
@expects_obj
def check(args):
"""
Check if the server configuration is working.
This command returns success if every checks pass,
or failure if any of these fails
"""
if args.nagios:
output.set_output_writer(output.NagiosOutputWriter())
servers = get_server_list(args)
for name in sorted(servers):
server = servers[name]
# Validate the returned server
if not manage_server_command(
server, name, skip_inactive=False,
skip_disabled=False, disabled_is_error=False):
continue
# If the server has been manually disabled
if not server.config.active:
name += " (inactive)"
# If server has configuration errors
elif server.config.disabled:
name += " (WARNING: disabled)"
output.init('check', name, server.config.active)
with closing(server):
server.check()
output.close_and_exit()
def diagnose():
"""
Diagnostic command (for support and problems detection purpose)
"""
# Get every server (both inactive and temporarily disabled)
servers = get_server_list(on_error_stop=False, suppress_error=True)
# errors list with duplicate paths between servers
errors_list = barman.__config__.servers_msg_list
barman.diagnose.exec_diagnose(servers, errors_list)
output.close_and_exit()
@named('show-backup')
@arg('server_name',
completer=server_completer,
help='specifies the server name for the command')
@arg('backup_id',
completer=backup_completer,
help='specifies the backup ID')
@expects_obj
def show_backup(args):
"""
This method shows a single backup information
"""
server = get_server(args)
# Retrieves the backup
backup_info = parse_backup_id(server, args)
with closing(server):
server.show_backup(backup_info)
output.close_and_exit()
@named('list-files')
@arg('server_name',
completer=server_completer,
help='specifies the server name for the command')
@arg('backup_id',
completer=backup_completer,
help='specifies the backup ID')
@arg('--target', choices=('standalone', 'data', 'wal', 'full'),
default='standalone',
help='''
Possible values are: data (just the data files), standalone
(base backup files, including required WAL files),
wal (just WAL files between the beginning of base
backup and the following one (if any) or the end of the log) and
full (same as data + wal). Defaults to %(default)s''')
@expects_obj
def list_files(args):
"""
List all the files for a single backup
"""
server = get_server(args)
# Retrieves the backup
backup_info = parse_backup_id(server, args)
try:
for line in backup_info.get_list_of_files(args.target):
output.info(line, log=False)
except BadXlogSegmentName as e:
output.error(
"invalid xlog segment name %r\n"
"HINT: Please run \"barman rebuild-xlogdb %s\" "
"to solve this issue",
str(e), server.config.name)
output.close_and_exit()
@arg('server_name',
completer=server_completer,
help='specifies the server name for the command')
@arg('backup_id',
completer=backup_completer,
help='specifies the backup ID')
@expects_obj
def delete(args):
"""
Delete a backup
"""
server = get_server(args)
# Retrieves the backup
backup_id = parse_backup_id(server, args)
with closing(server):
server.delete_backup(backup_id)
output.close_and_exit()
@named('get-wal')
@arg('server_name',
completer=server_completer,
help='specifies the server name for the command')
@arg('wal_name',
help='the WAL file to get')
@arg('--output-directory', '-o',
help='put the retrieved WAL file in this directory '
'with the original name',
default=SUPPRESS)
@arg('--gzip', '-z', '-x',
help='compress the output with gzip',
action='store_const', const='gzip', dest='compression', default=SUPPRESS)
@arg('--bzip2', '-j',
help='compress the output with bzip2',
action='store_const', const='bzip2', dest='compression', default=SUPPRESS)
@arg('--peek', '-p',
help="peek from the WAL archive up to 'SIZE' WAL files, starting "
"from the requested one. 'SIZE' must be an integer >= 1. "
"When invoked with this option, get-wal returns a list of "
"zero to 'SIZE' WAL segment names, one per row.",
metavar='SIZE',
type=check_positive,
default=SUPPRESS)
@expects_obj
def get_wal(args):
"""
Retrieve WAL_NAME file from SERVER_NAME archive.
The content will be streamed on standard output unless
the --output-directory option is specified.
"""
server = get_server(args, inactive_is_error=True)
# Retrieve optional arguments. If an argument is not specified,
# the namespace doesn't contain it due to SUPPRESS default.
# In that case we pick 'None' using getattr third argument.
compression = getattr(args, 'compression', None)
output_directory = getattr(args, 'output_directory', None)
peek = getattr(args, 'peek', None)
with closing(server):
server.get_wal(args.wal_name,
compression=compression,
output_directory=output_directory,
peek=peek)
output.close_and_exit()
@named('archive-wal')
@arg('server_name',
completer=server_completer,
help='specifies the server name for the command')
@expects_obj
def archive_wal(args):
"""
Execute maintenance operations on WAL files for a given server.
This command processes any incoming WAL files for the server
and archives them along the catalogue.
"""
server = get_server(args)
with closing(server):
server.archive_wal()
output.close_and_exit()
@named('receive-wal')
@arg('--stop', help='stop the receive-wal subprocess for the server',
action='store_true')
@arg('--reset', help='reset the status of receive-wal removing '
'any status files',
action='store_true')
@arg('--create-slot', help='create the replication slot, if it does not exist',
action='store_true')
@arg('--drop-slot', help='drop the replication slot, if it exists',
action='store_true')
@arg('server_name',
completer=server_completer,
help='specifies the server name for the command')
@expects_obj
def receive_wal(args):
"""
Start a receive-wal process.
The process uses the streaming protocol to receive WAL files
from the PostgreSQL server.
"""
server = get_server(args)
if args.stop and args.reset:
output.error("--stop and --reset options are not compatible")
# If the caller requested to shutdown the receive-wal process deliver the
# termination signal, otherwise attempt to start it
elif args.stop:
server.kill('receive-wal')
elif args.create_slot:
with closing(server):
server.create_physical_repslot()
elif args.drop_slot:
with closing(server):
server.drop_repslot()
else:
with closing(server):
server.receive_wal(reset=args.reset)
output.close_and_exit()
def pretty_args(args):
"""
Prettify the given argh namespace to be human readable
:type args: argh.dispatching.ArghNamespace
:return: the human readable content of the namespace
"""
values = dict(vars(args))
# Retrieve the command name with recent argh versions
if '_functions_stack' in values:
values['command'] = values['_functions_stack'][0].__name__
del values['_functions_stack']
# Older argh versions only have the matching function in the namespace
elif 'function' in values:
values['command'] = values['function'].__name__
del values['function']
return "%r" % values
def global_config(args):
"""
Set the configuration file
"""
if hasattr(args, 'config'):
filename = args.config
else:
try:
filename = os.environ['BARMAN_CONFIG_FILE']
except KeyError:
filename = None
config = barman.config.Config(filename)
barman.__config__ = config
# change user if needed
try:
drop_privileges(config.user)
except OSError:
msg = "ERROR: please run barman as %r user" % config.user
raise SystemExit(msg)
except KeyError:
msg = "ERROR: the configured user %r does not exists" % config.user
raise SystemExit(msg)
# configure logging
log_level = parse_log_level(config.log_level)
configure_logging(config.log_file,
log_level or barman.config.DEFAULT_LOG_LEVEL,
config.log_format)
if log_level is None:
_logger.warn('unknown log_level in config file: %s', config.log_level)
# configure output
if args.format != output.DEFAULT_WRITER or args.quiet or args.debug:
output.set_output_writer(args.format,
quiet=args.quiet,
debug=args.debug)
# Load additional configuration files
config.load_configuration_files_directory()
# We must validate the configuration here in order to have
# both output and logging configured
config.validate_global_config()
_logger.debug('Initialised Barman version %s (config: %s, args: %s)',
barman.__version__, config.config_file, pretty_args(args))
def get_server(args, skip_inactive=True, skip_disabled=False,
inactive_is_error=False,
on_error_stop=True, suppress_error=False):
"""
Get a single server retrieving its configuration (wraps get_server_list())
Returns a Server object or None if the required server is unknown and
on_error_stop is False.
WARNING: this function modifies the 'args' parameter
:param args: an argparse namespace containing a single
server_name parameter
WARNING: the function modifies the content of this parameter
:param bool skip_inactive: do nothing if the server is inactive
:param bool skip_disabled: do nothing if the server is disabled
:param bool inactive_is_error: treat inactive server as error
:param bool on_error_stop: stop if an error is found
:param bool suppress_error: suppress display of errors (e.g. diagnose)
:rtype: barman.server.Server|None
"""
# This function must to be called with in a single-server context
name = args.server_name
assert isinstance(name, str)
# The 'all' special name is forbidden in this context
if name == 'all':
output.error("You cannot use 'all' in a single server context")
output.close_and_exit()
# The following return statement will never be reached
# but it is here for clarity
return None
# Builds a list from a single given name
args.server_name = [name]
# Skip_inactive is reset if inactive_is_error is set, because
# it needs to retrieve the inactive server to emit the error.
skip_inactive &= not inactive_is_error
# Retrieve the requested server
servers = get_server_list(args, skip_inactive, skip_disabled,
on_error_stop, suppress_error)
# The requested server has been excluded from get_server_list result
if len(servers) == 0:
output.close_and_exit()
# The following return statement will never be reached
# but it is here for clarity
return None
# retrieve the server object
server = servers[name]
# Apply standard validation control and skips
# the server if inactive or disabled, displaying standard
# error messages. If on_error_stop (default) exits
if not manage_server_command(server, name,
inactive_is_error) and \
on_error_stop:
output.close_and_exit()
# The following return statement will never be reached
# but it is here for clarity
return None
# Returns the filtered server
return server
def get_server_list(args=None, skip_inactive=False, skip_disabled=False,
on_error_stop=True, suppress_error=False):
"""
Get the server list from the configuration
If args the parameter is None or arg.server_name is ['all']
returns all defined servers
:param args: an argparse namespace containing a list server_name parameter
:param bool skip_inactive: skip inactive servers when 'all' is required
:param bool skip_disabled: skip disabled servers when 'all' is required
:param bool on_error_stop: stop if an error is found
:param bool suppress_error: suppress display of errors (e.g. diagnose)
:rtype: dict(str,barman.server.Server|None)
"""
server_dict = {}
# This function must to be called with in a multiple-server context
assert not args or isinstance(args.server_name, list)
# Generate the list of servers (required for global errors)
available_servers = barman.__config__.server_names()
# Get a list of configuration errors from all the servers
global_error_list = barman.__config__.servers_msg_list
# Global errors have higher priority
if global_error_list:
# Output the list of global errors
if not suppress_error:
for error in global_error_list:
output.error(error)
# If requested, exit on first error
if on_error_stop:
output.close_and_exit()
# The following return statement will never be reached
# but it is here for clarity
return {}
# Handle special 'all' server cases
# - args is None
# - 'all' special name
if not args or 'all' in args.server_name:
# When 'all' is used, it must be the only specified argument
if args and len(args.server_name) != 1:
output.error("You cannot use 'all' with other server names")
servers = available_servers
else:
servers = args.server_name
# Loop through all the requested servers
for server in servers:
conf = barman.__config__.get_server(server)
if conf is None:
# Unknown server
server_dict[server] = None
else:
server_object = Server(conf)
# Skip inactive servers, if requested
if skip_inactive and not server_object.config.active:
output.info("Skipping inactive server '%s'"
% conf.name)
continue
# Skip disabled servers, if requested
if skip_disabled and server_object.config.disabled:
output.info("Skipping temporarily disabled server '%s'"
% conf.name)
continue
server_dict[server] = server_object
return server_dict
def manage_server_command(server,
name=None,
inactive_is_error=False,
disabled_is_error=True,
skip_inactive=True,
skip_disabled=True):
"""
Standard and consistent method for managing server errors within
a server command execution. By default, suggests to skip any inactive
and disabled server; it also emits errors for disabled servers by
default.
Returns True if the command has to be executed for this server.
:param barman.server.Server server: server to be checked for errors
:param str name: name of the server, in a multi-server command
:param bool inactive_is_error: treat inactive server as error
:param bool disabled_is_error: treat disabled server as error
:param bool skip_inactive: skip if inactive
:param bool skip_disabled: skip if disabled
:return: True if the command has to be executed on this server
:rtype: boolean
"""
# Unknown server (skip it)
if not server:
output.error("Unknown server '%s'" % name)
return False
if not server.config.active:
# Report inactive server as error
if inactive_is_error:
output.error('Inactive server: %s' % server.config.name)
if skip_inactive:
return False
# Report disabled server as error
if server.config.disabled:
# Output all the messages as errors, and exit terminating the run.
if disabled_is_error:
for message in server.config.msg_list:
output.error(message)
if skip_disabled:
return False
# All ok, execute the command
return True
def parse_backup_id(server, args):
"""
Parses backup IDs including special words such as latest, oldest, etc.
Exit with error if the backup id doesn't exist.
:param Server server: server object to search for the required backup
:param args: command lien arguments namespace
:rtype: BackupInfo
"""
if args.backup_id in ('latest', 'last'):
backup_id = server.get_last_backup_id()
elif args.backup_id in ('oldest', 'first'):
backup_id = server.get_first_backup_id()
else:
backup_id = args.backup_id
backup_info = server.get_backup(backup_id)
if backup_info is None:
output.error(
"Unknown backup '%s' for server '%s'",
args.backup_id, server.config.name)
output.close_and_exit()
return backup_info
def main():
"""
The main method of Barman
"""
p = ArghParser(epilog='Barman by 2ndQuadrant (www.2ndQuadrant.com)')
p.add_argument('-v', '--version', action='version',
version='%s\n\nBarman by 2ndQuadrant (www.2ndQuadrant.com)'
% barman.__version__)
p.add_argument('-c', '--config',
help='uses a configuration file '
'(defaults: %s)'
% ', '.join(barman.config.Config.CONFIG_FILES),
default=SUPPRESS)
p.add_argument('-q', '--quiet', help='be quiet', action='store_true')
p.add_argument('-d', '--debug', help='debug output', action='store_true')
p.add_argument('-f', '--format', help='output format',
choices=output.AVAILABLE_WRITERS.keys(),
default=output.DEFAULT_WRITER)
p.add_commands(
[
archive_wal,
backup,
check,
cron,
delete,
diagnose,
get_wal,
list_backup,
list_files,
list_server,
rebuild_xlogdb,
receive_wal,
recover,
show_backup,
show_server,
replication_status,
status,
switch_xlog,
]
)
# noinspection PyBroadException
try:
p.dispatch(pre_call=global_config)
except KeyboardInterrupt:
msg = "Process interrupted by user (KeyboardInterrupt)"
output.error(msg)
except Exception as e:
msg = "%s\nSee log file for more details." % e
output.exception(msg)
# cleanup output API and exit honoring output.error_occurred and
# output.error_exit_code
output.close_and_exit()
if __name__ == '__main__':
# This code requires the mock module and allow us to test
# bash completion inside the IDE debugger
try:
# noinspection PyUnresolvedReferences
import mock
sys.stdout = mock.Mock(wraps=sys.stdout)
sys.stdout.isatty.return_value = True
os.dup2(2, 8)
except ImportError:
pass
main()
barman-2.1/barman/remote_status.py 0000664 0001750 0001750 00000004363 12726221047 017733 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
Remote Status module
A Remote Status class implements a standard interface for
retrieving and caching the results of a remote component
(such as Postgres server, WAL archiver, etc.). It follows
the Mixin pattern.
"""
from abc import ABCMeta, abstractmethod
from barman.utils import with_metaclass
class RemoteStatusMixin(with_metaclass(ABCMeta, object)):
"""
Abstract base class that implements remote status capabilities
following the Mixin pattern.
"""
def __init__(self, *args, **kwargs):
"""
Base constructor (Mixin pattern)
"""
self._remote_status = None
super(RemoteStatusMixin, self).__init__(*args, **kwargs)
@abstractmethod
def fetch_remote_status(self):
"""
Retrieve status information from the remote component
The implementation of this method must not raise any exception in case
of errors, but should set the missing values to None in the resulting
dictionary.
:rtype: dict[str, None|str]
"""
def get_remote_status(self):
"""
Get the status of the remote component
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
if self._remote_status is None:
self._remote_status = self.fetch_remote_status()
return self._remote_status
def reset_remote_status(self):
"""
Reset the cached result
"""
self._remote_status = None
barman-2.1/barman/utils.py 0000664 0001750 0001750 00000026001 13026732452 016167 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module contains utility functions used in Barman.
"""
import datetime
import decimal
import errno
import grp
import json
import logging
import logging.handlers
import os
import pwd
import re
import signal
from contextlib import contextmanager
from distutils.version import Version
from barman.exceptions import TimeoutError
_logger = logging.getLogger(__name__)
def drop_privileges(user):
"""
Change the system user of the current python process.
It will only work if called as root or as the target user.
:param string user: target user
:raise KeyError: if the target user doesn't exists
:raise OSError: when the user change fails
"""
pw = pwd.getpwnam(user)
if pw.pw_uid == os.getuid():
return
groups = [e.gr_gid for e in grp.getgrall() if pw.pw_name in e.gr_mem]
groups.append(pw.pw_gid)
os.setgroups(groups)
os.setgid(pw.pw_gid)
os.setuid(pw.pw_uid)
os.environ['HOME'] = pw.pw_dir
def mkpath(directory):
"""
Recursively create a target directory.
If the path already exists it does nothing.
:param str directory: directory to be created
"""
if not os.path.isdir(directory):
os.makedirs(directory)
def configure_logging(
log_file,
log_level=logging.INFO,
log_format="%(asctime)s %(name)s %(levelname)s: %(message)s"):
"""
Configure the logging module
:param str,None log_file: target file path. If None use standard error.
:param int log_level: min log level to be reported in log file.
Default to INFO
:param str log_format: format string used for a log line.
Default to "%(asctime)s %(name)s %(levelname)s: %(message)s"
"""
warn = None
handler = logging.StreamHandler()
if log_file:
log_file = os.path.abspath(log_file)
log_dir = os.path.dirname(log_file)
try:
mkpath(log_dir)
handler = logging.handlers.WatchedFileHandler(log_file)
except (OSError, IOError):
# fallback to standard error
warn = "Failed opening the requested log file. " \
"Using standard error instead."
formatter = logging.Formatter(log_format)
handler.setFormatter(formatter)
logging.root.addHandler(handler)
if warn:
# this will be always displayed because the default level is WARNING
_logger.warn(warn)
logging.root.setLevel(log_level)
def parse_log_level(log_level):
"""
Convert a log level to its int representation as required by
logging module.
:param log_level: An integer or a string
:return: an integer or None if an invalid argument is provided
"""
try:
log_level_int = int(log_level)
except ValueError:
log_level_int = logging.getLevelName(str(log_level).upper())
if isinstance(log_level_int, int):
return log_level_int
return None
def pretty_size(size, unit=1024):
"""
This function returns a pretty representation of a size value
:param int|long|float size: the number to to prettify
:param int unit: 1000 or 1024 (the default)
:rtype: str
"""
suffixes = ["B"] + [i + {1000: "B", 1024: "iB"}[unit] for i in "KMGTPEZY"]
if unit == 1000:
suffixes[1] = 'kB' # special case kB instead of KB
# cast to float to avoid loosing decimals
size = float(size)
for suffix in suffixes:
if abs(size) < unit or suffix == suffixes[-1]:
if suffix == suffixes[0]:
return "%d %s" % (size, suffix)
else:
return "%.1f %s" % (size, suffix)
else:
size /= unit
def human_readable_timedelta(timedelta):
"""
Given a time interval, returns a human readable string
:param timedelta: the timedelta to transform in a human readable form
"""
delta = abs(timedelta)
# Calculate time units for the given interval
time_map = {
'day': int(delta.days),
'hour': int(delta.seconds / 3600),
'minute': int(delta.seconds / 60) % 60,
'second': int(delta.seconds % 60),
}
# Build the resulting string
time_list = []
# 'Day' part
if time_map['day'] > 0:
if time_map['day'] == 1:
time_list.append('%s day' % time_map['day'])
else:
time_list.append('%s days' % time_map['day'])
# 'Hour' part
if time_map['hour'] > 0:
if time_map['hour'] == 1:
time_list.append('%s hour' % time_map['hour'])
else:
time_list.append('%s hours' % time_map['hour'])
# 'Minute' part
if time_map['minute'] > 0:
if time_map['minute'] == 1:
time_list.append('%s minute' % time_map['minute'])
else:
time_list.append('%s minutes' % time_map['minute'])
# 'Second' part
if time_map['second'] > 0:
if time_map['second'] == 1:
time_list.append('%s second' % time_map['second'])
else:
time_list.append('%s seconds' % time_map['second'])
human = ', '.join(time_list)
# Take care of timedelta when is shorter than a second
if delta < datetime.timedelta(seconds=1):
human = 'less than one second'
# If timedelta is negative append 'ago' suffix
if delta != timedelta:
human += " ago"
return human
def which(executable, path=None):
"""
This method is useful to find if a executable is present into the
os PATH
:param str executable: The name of the executable to find
:param str|None path: An optional search path to override the current one.
:return str|None: the path of the executable or None
"""
# Get the system path and split.
if path is None:
path = os.getenv('PATH')
# If executable is an absolute path, check if it exists and is executable
# otherwise return failure.
if os.path.isabs(executable):
if os.path.exists(executable) and os.access(executable, os.X_OK):
return executable
else:
return None
# Search the requested executable in eery directory present in path and
# return the first occurrence that exists and is executable.
for file_path in path.split(os.path.pathsep):
file_path = os.path.join(file_path, executable)
# If the file exists and is executable return the full path.
if os.path.exists(file_path) and os.access(file_path, os.X_OK):
return file_path
# If no matching file is present on the system return None
return None
class BarmanEncoder(json.JSONEncoder):
"""
Custom JSON encoder used for BackupInfo encoding
This encoder supports the following types:
* dates and timestamps if they have a ctime() method.
* objects that implement the 'to_json' method.
* binary strings (python 3)
"""
def default(self, obj):
# If the object implements to_json() method use it
if hasattr(obj, 'to_json'):
return obj.to_json()
# Serialise date and datetime objects using ctime() method
if hasattr(obj, 'ctime') and callable(obj.ctime):
return obj.ctime()
# Serialise timedelta objects using human_readable_timedelta()
if isinstance(obj, datetime.timedelta):
return human_readable_timedelta(obj)
# Serialise Decimal objects using their string representation
# WARNING: When deserialized they will be treat as float values
# which have a lower precision
if isinstance(obj, decimal.Decimal):
return float(obj)
# Binary strings must be decoded before using them in
# an unicode string
if hasattr(obj, 'decode') and callable(obj.decode):
return obj.decode('utf-8', 'replace')
# Manage (Loose|Strict)Version objects as strings.
if isinstance(obj, Version):
return str(obj)
# Let the base class default method raise the TypeError
return super(BarmanEncoder, self).default(obj)
def fsync_dir(dir_path):
"""
Execute fsync on a directory ensuring it is synced to disk
:param str dir_path: The directory to sync
:raise OSError: If fail opening the directory
"""
dir_fd = os.open(dir_path, os.O_DIRECTORY)
try:
os.fsync(dir_fd)
except OSError as e:
# On some filesystem doing a fsync on a directory
# raises an EINVAL error. Ignoring it is usually safe.
if e.errno != errno.EINVAL:
raise
os.close(dir_fd)
def simplify_version(version_string):
"""
Simplify a version number by removing the patch level
:param version_string: the version number to simplify
:return str: the simplified version number
"""
if version_string is None:
return None
version = version_string.split('.')
# If a development/beta/rc version, split out the string part
unreleased = re.search(r'[^0-9.]', version[-1])
if unreleased:
last_component = version.pop()
number = last_component[:unreleased.start()]
string = last_component[unreleased.start():]
version += [number, string]
return '.'.join(version[:-1])
def with_metaclass(meta, *bases):
"""
Function from jinja2/_compat.py. License: BSD.
Create a base class with a metaclass.
:param type meta: Metaclass to add to base class
"""
# This requires a bit of explanation: the basic idea is to make a
# dummy metaclass for one level of class instantiation that replaces
# itself with the actual metaclass.
class Metaclass(type):
def __new__(mcs, name, this_bases, d):
return meta(name, bases, d)
return type.__new__(Metaclass, 'temporary_class', (), {})
@contextmanager
def timeout(timeout_duration):
"""
ContextManager responsible for timing out the contained
block of code after a defined time interval.
"""
# Define the handler for the alarm signal
def handler(signum, frame):
raise TimeoutError()
# set the timeout handler
previous_handler = signal.signal(signal.SIGALRM, handler)
if previous_handler != signal.SIG_DFL:
signal.signal(signal.SIGALRM, previous_handler)
raise AssertionError("Another timeout is already defined")
# set the timeout duration
signal.alarm(timeout_duration)
try:
# Execute the contained block of code
yield
finally:
# Reset the signal
signal.alarm(0)
signal.signal(signal.SIGALRM, signal.SIG_DFL)
barman-2.1/barman/config.py 0000664 0001750 0001750 00000073620 13033425127 016301 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module is responsible for all the things related to
Barman configuration, such as parsing configuration file.
"""
import collections
import datetime
import inspect
import logging.handlers
import os
import re
import sys
from glob import iglob
from barman import output
try:
from ConfigParser import ConfigParser, NoOptionError
except ImportError:
from configparser import ConfigParser, NoOptionError
# create a namedtuple object called PathConflict with 'label' and 'server'
PathConflict = collections.namedtuple('PathConflict', 'label server')
_logger = logging.getLogger(__name__)
FORBIDDEN_SERVER_NAMES = ['all']
DEFAULT_USER = 'barman'
DEFAULT_LOG_LEVEL = logging.INFO
DEFAULT_LOG_FORMAT = "%(asctime)s [%(process)s] %(name)s " \
"%(levelname)s: %(message)s"
_TRUE_RE = re.compile(r"""^(true|t|yes|1|on)$""", re.IGNORECASE)
_FALSE_RE = re.compile(r"""^(false|f|no|0|off)$""", re.IGNORECASE)
_TIME_INTERVAL_RE = re.compile(r"""
^\s*
(\d+)\s+(day|month|week)s? # N (day|month|week) with optional 's'
\s*$
""", re.IGNORECASE | re.VERBOSE)
REUSE_BACKUP_VALUES = ('copy', 'link', 'off')
# Possible copy methods for backups (must be all lowercase)
BACKUP_METHOD_VALUES = ['rsync', 'postgres']
class CsvOption(set):
"""
Base class for CSV options.
Given a comma delimited string, this class is a list containing the
submitted options.
Internally, it uses a set in order to avoid option replication.
Allowed values for the CSV option are contained in the 'value_list'
attribute.
The 'conflicts' attribute specifies for any value, the list of
values that are prohibited (and thus generate a conflict).
If a conflict is found, raises a ValueError exception.
"""
value_list = []
conflicts = {}
def __init__(self, value, key, source):
# Invoke parent class init and initialize an empty set
super(CsvOption, self).__init__()
# Parse not None values
if value is not None:
self.parse(value, key, source)
# Validates the object structure before returning the new instance
self.validate(key, source)
def parse(self, value, key, source):
"""
Parses a list of values and correctly assign the set of values
(removing duplication) and checking for conflicts.
"""
if not value:
return
values_list = value.split(',')
for val in sorted(values_list):
val = val.strip().lower()
if val in self.value_list:
# check for conflicting values. if a conflict is
# found the option is not valid then, raise exception.
if val in self.conflicts and self.conflicts[val] in self:
raise ValueError("Invalid configuration value '%s' for "
"key %s in %s: cannot contain both "
"'%s' and '%s'."
"Configuration directive ignored." %
(val, key, source, val,
self.conflicts[val]))
else:
# otherwise use parsed value
self.add(val)
else:
# not allowed value, reject the configuration
raise ValueError("Invalid configuration value '%s' for "
"key %s in %s: Unknown option" %
(val, key, source))
def validate(self, key, source):
"""
Override this method for special validation needs
"""
def to_json(self):
"""
Output representation of the obj for JSON serialization
The result is a string which can be parsed by the same class
"""
return ",".join(self)
class BackupOptions(CsvOption):
"""
Extends CsvOption class providing all the details for the backup_options
field
"""
# constants containing labels for allowed values
EXCLUSIVE_BACKUP = 'exclusive_backup'
CONCURRENT_BACKUP = 'concurrent_backup'
# list holding all the allowed values for the BackupOption class
value_list = [EXCLUSIVE_BACKUP, CONCURRENT_BACKUP]
# map holding all the possible conflicts between the allowed values
conflicts = {
EXCLUSIVE_BACKUP: CONCURRENT_BACKUP,
CONCURRENT_BACKUP: EXCLUSIVE_BACKUP, }
def validate(self, key, source):
"""
Validates backup_option values: currently it makes sure
that either exclusive_backup or concurrent_backup are set.
"""
if len(self) == 0:
return
if self.CONCURRENT_BACKUP not in self \
and self.EXCLUSIVE_BACKUP not in self:
raise ValueError("Invalid configuration value for "
"key %s in %s: it must contain either "
"exclusive_backup or concurrent_backup option"
% (key, source))
class RecoveryOptions(CsvOption):
"""
Extends CsvOption class providing all the details for the recovery_options
field
"""
# constants containing labels for allowed values
GET_WAL = 'get-wal'
# list holding all the allowed values for the RecoveryOptions class
value_list = [GET_WAL]
def parse_boolean(value):
"""
Parse a string to a boolean value
:param str value: string representing a boolean
:raises ValueError: if the string is an invalid boolean representation
"""
if _TRUE_RE.match(value):
return True
if _FALSE_RE.match(value):
return False
raise ValueError("Invalid boolean representation (use 'true' or 'false')")
def parse_time_interval(value):
"""
Parse a string, transforming it in a time interval.
Accepted format: N (day|month|week)s
:param str value: the string to evaluate
"""
# if empty string or none return none
if value is None or value == '':
return None
result = _TIME_INTERVAL_RE.match(value)
# if the string doesn't match, the option is invalid
if not result:
raise ValueError("Invalid value for a time interval %s" %
value)
# if the int conversion
value = int(result.groups()[0])
unit = result.groups()[1][0].lower()
# Calculates the time delta
if unit == 'd':
time_delta = datetime.timedelta(days=value)
elif unit == 'w':
time_delta = datetime.timedelta(weeks=value)
elif unit == 'm':
time_delta = datetime.timedelta(days=(31 * value))
else:
# This should never happen
raise ValueError("Invalid unit time %s" % unit)
return time_delta
def parse_reuse_backup(value):
"""
Parse a string to a valid reuse_backup value.
Valid values are "copy", "link" and "off"
:param str value: reuse_backup value
:raises ValueError: if the value is invalid
"""
if value is None:
return None
if value.lower() in REUSE_BACKUP_VALUES:
return value.lower()
raise ValueError(
"Invalid value (use '%s' or '%s')" % (
"', '".join(REUSE_BACKUP_VALUES[:-1]), REUSE_BACKUP_VALUES[-1]))
def parse_backup_method(value):
"""
Parse a string to a valid backup_method value.
Valid values are contained in BACKUP_METHOD_VALUES list
:param str value: backup_method value
:raises ValueError: if the value is invalid
"""
if value is None:
return None
if value.lower() in BACKUP_METHOD_VALUES:
return value.lower()
raise ValueError(
"Invalid value (must be one in: '%s')" % (
"', '".join(BACKUP_METHOD_VALUES)))
class ServerConfig(object):
"""
This class represents the configuration for a specific Server instance.
"""
KEYS = [
'active',
'archiver',
'archiver_batch_size',
'backup_directory',
'backup_method',
'backup_options',
'bandwidth_limit',
'basebackup_retry_sleep',
'basebackup_retry_times',
'basebackups_directory',
'check_timeout',
'compression',
'conninfo',
'custom_compression_filter',
'custom_decompression_filter',
'description',
'disabled',
'errors_directory',
'immediate_checkpoint',
'incoming_wals_directory',
'last_backup_maximum_age',
'minimum_redundancy',
'network_compression',
'path_prefix',
'post_archive_retry_script',
'post_archive_script',
'post_backup_retry_script',
'post_backup_script',
'pre_archive_retry_script',
'pre_archive_script',
'pre_backup_retry_script',
'pre_backup_script',
'recovery_options',
'retention_policy',
'retention_policy_mode',
'reuse_backup',
'slot_name',
'ssh_command',
'streaming_archiver',
'streaming_archiver_batch_size',
'streaming_archiver_name',
'streaming_backup_name',
'streaming_conninfo',
'streaming_wals_directory',
'tablespace_bandwidth_limit',
'wal_retention_policy',
'wals_directory'
]
BARMAN_KEYS = [
'archiver',
'archiver_batch_size',
'backup_method',
'backup_options',
'bandwidth_limit',
'basebackup_retry_sleep',
'basebackup_retry_times',
'check_timeout',
'compression',
'configuration_files_directory',
'custom_compression_filter',
'custom_decompression_filter',
'immediate_checkpoint',
'last_backup_maximum_age',
'minimum_redundancy',
'network_compression',
'path_prefix',
'post_archive_retry_script',
'post_archive_script',
'post_backup_retry_script',
'post_backup_script',
'pre_archive_retry_script',
'pre_archive_script',
'pre_backup_retry_script',
'pre_backup_script',
'recovery_options',
'retention_policy',
'retention_policy_mode',
'reuse_backup',
'slot_name',
'streaming_archiver',
'streaming_archiver_batch_size',
'streaming_archiver_name',
'streaming_backup_name',
'tablespace_bandwidth_limit',
'wal_retention_policy'
]
DEFAULTS = {
'active': 'true',
'archiver': 'off',
'archiver_batch_size': '0',
'backup_directory': '%(barman_home)s/%(name)s',
'backup_method': 'rsync',
'backup_options': '',
'basebackup_retry_sleep': '30',
'basebackup_retry_times': '0',
'basebackups_directory': '%(backup_directory)s/base',
'check_timeout': '30',
'disabled': 'false',
'errors_directory': '%(backup_directory)s/errors',
'immediate_checkpoint': 'false',
'incoming_wals_directory': '%(backup_directory)s/incoming',
'minimum_redundancy': '0',
'network_compression': 'false',
'recovery_options': '',
'retention_policy_mode': 'auto',
'streaming_archiver': 'off',
'streaming_archiver_batch_size': '0',
'streaming_archiver_name': 'barman_receive_wal',
'streaming_backup_name': 'barman_streaming_backup',
'streaming_conninfo': '%(conninfo)s',
'streaming_wals_directory': '%(backup_directory)s/streaming',
'wal_retention_policy': 'main',
'wals_directory': '%(backup_directory)s/wals'
}
FIXED = [
'disabled',
]
PARSERS = {
'active': parse_boolean,
'archiver': parse_boolean,
'archiver_batch_size': int,
'backup_method': parse_backup_method,
'backup_options': BackupOptions,
'basebackup_retry_sleep': int,
'basebackup_retry_times': int,
'check_timeout': int,
'disabled': parse_boolean,
'immediate_checkpoint': parse_boolean,
'last_backup_maximum_age': parse_time_interval,
'network_compression': parse_boolean,
'recovery_options': RecoveryOptions,
'reuse_backup': parse_reuse_backup,
'streaming_archiver': parse_boolean,
'streaming_archiver_batch_size': int,
}
def invoke_parser(self, key, source, value, new_value):
"""
Function used for parsing configuration values.
If needed, it uses special parsers from the PARSERS map,
and handles parsing exceptions.
Uses two values (value and new_value) to manage
configuration hierarchy (server config overwrites global config).
:param str key: the name of the configuration option
:param str source: the section that contains the configuration option
:param value: the old value of the option if present.
:param str new_value: the new value that needs to be parsed
:return: the parsed value of a configuration option
"""
# If the new value is None, returns the old value
if new_value is None:
return value
# If we have a parser for the current key, use it to obtain the
# actual value. If an exception is thrown, print a warning and
# ignore the value.
# noinspection PyBroadException
if key in self.PARSERS:
parser = self.PARSERS[key]
try:
# If the parser is a subclass of the CsvOption class
# we need a different invocation, which passes not only
# the value to the parser, but also the key name
# and the section that contains the configuration
if inspect.isclass(parser) \
and issubclass(parser, CsvOption):
value = parser(new_value, key, source)
else:
value = parser(new_value)
except Exception as e:
output.warning("Ignoring invalid configuration value '%s' "
"for key %s in %s: %s",
new_value, key, source, e)
else:
value = new_value
return value
def __init__(self, config, name):
self.msg_list = []
self.config = config
self.name = name
self.barman_home = config.barman_home
self.barman_lock_directory = config.barman_lock_directory
config.validate_server_config(self.name)
for key in ServerConfig.KEYS:
value = None
# Skip parameters that cannot be configured by users
if key not in ServerConfig.FIXED:
# Get the setting from the [name] section of config file
# A literal None value is converted to an empty string
new_value = config.get(name, key, self.__dict__, none_value='')
source = '[%s] section' % name
value = self.invoke_parser(key, source, value, new_value)
# If the setting isn't present in [name] section of config file
# check if it has to be inherited from the [barman] section
if value is None and key in ServerConfig.BARMAN_KEYS:
new_value = config.get('barman',
key,
self.__dict__,
none_value='')
source = '[barman] section'
value = self.invoke_parser(key, source, value, new_value)
# If the setting isn't present in [name] section of config file
# and is not inherited from global section use its default
# (if present)
if value is None and key in ServerConfig.DEFAULTS:
new_value = ServerConfig.DEFAULTS[key] % self.__dict__
source = 'DEFAULTS'
value = self.invoke_parser(key, source, value, new_value)
# An empty string is a None value (bypassing inheritance
# from global configuration)
if value is not None and value == '' or value == 'None':
value = None
setattr(self, key, value)
def to_json(self):
"""
Return an equivalent dictionary that can be encoded in json
"""
json_dict = dict(vars(self))
# remove the reference to main Config object
del json_dict['config']
return json_dict
def get_bwlimit(self, tablespace=None):
"""
Return the configured bandwidth limit for the provided tablespace
If tablespace is None, it returns the global bandwidth limit
:param barman.infofile.Tablespace tablespace: the tablespace to copy
:rtype: str
"""
# Default to global bandwidth limit
bwlimit = self.bandwidth_limit
if tablespace:
# A tablespace can be copied using a per-tablespace bwlimit
tablespaces_bw_limit = self.tablespace_bandwidth_limit
if (tablespaces_bw_limit and
tablespace.name in tablespaces_bw_limit):
bwlimit = tablespaces_bw_limit[tablespace.name]
return bwlimit
class Config(object):
"""This class represents the barman configuration.
Default configuration files are /etc/barman.conf,
/etc/barman/barman.conf
and ~/.barman.conf for a per-user configuration
"""
CONFIG_FILES = [
'~/.barman.conf',
'/etc/barman.conf',
'/etc/barman/barman.conf',
]
_QUOTE_RE = re.compile(r"""^(["'])(.*)\1$""")
def __init__(self, filename=None):
self._config = ConfigParser()
if filename:
if hasattr(filename, 'read'):
self._config.readfp(filename)
else:
# check for the existence of the user defined file
if not os.path.exists(filename):
sys.exit("Configuration file '%s' does not exist" %
filename)
self._config.read(os.path.expanduser(filename))
else:
# Check for the presence of configuration files
# inside default directories
for path in self.CONFIG_FILES:
full_path = os.path.expanduser(path)
if os.path.exists(full_path) \
and full_path in self._config.read(full_path):
filename = full_path
break
else:
sys.exit("Could not find any configuration file at "
"default locations.\n"
"Check Barman's documentation for more help.")
self.config_file = filename
self._servers = None
self.servers_msg_list = []
self._parse_global_config()
def get(self, section, option, defaults=None, none_value=None):
"""Method to get the value from a given section from
Barman configuration
"""
if not self._config.has_section(section):
return None
try:
value = self._config.get(section, option, raw=False, vars=defaults)
if value.lower() == 'none':
value = none_value
if value is not None:
value = self._QUOTE_RE.sub(lambda m: m.group(2), value)
return value
except NoOptionError:
return None
def _parse_global_config(self):
"""
This method parses the global [barman] section
"""
self.barman_home = self.get('barman', 'barman_home')
self.barman_lock_directory = self.get(
'barman', 'barman_lock_directory') or self.barman_home
self.user = self.get('barman', 'barman_user') or DEFAULT_USER
self.log_file = self.get('barman', 'log_file')
self.log_format = self.get(
'barman', 'log_format') or DEFAULT_LOG_FORMAT
self.log_level = self.get('barman', 'log_level') or DEFAULT_LOG_LEVEL
# save the raw barman section to be compared later in
# _is_global_config_changed() method
self._global_config = set(self._config.items('barman'))
def _is_global_config_changed(self):
"""Return true if something has changed in global configuration"""
return self._global_config != set(self._config.items('barman'))
def load_configuration_files_directory(self):
"""
Read the "configuration_files_directory" option and load all the
configuration files with the .conf suffix that lie in that folder
"""
config_files_directory = self.get('barman',
'configuration_files_directory')
if not config_files_directory:
return
if not os.path.isdir(os.path.expanduser(config_files_directory)):
_logger.warn(
'Ignoring the "configuration_files_directory" option as "%s" '
'is not a directory',
config_files_directory)
return
for cfile in sorted(iglob(
os.path.join(os.path.expanduser(config_files_directory),
'*.conf'))):
filename = os.path.basename(cfile)
if os.path.isfile(cfile):
# Load a file
_logger.debug('Including configuration file: %s', filename)
self._config.read(cfile)
if self._is_global_config_changed():
msg = "the configuration file %s contains a not empty [" \
"barman] section" % filename
_logger.fatal(msg)
raise SystemExit("FATAL: %s" % msg)
else:
# Add an info that a file has been discarded
_logger.warn('Discarding configuration file: %s (not a file)',
filename)
def _populate_servers(self):
"""
Populate server list from configuration file
Also check for paths errors in configuration.
If two or more paths overlap in
a single server, that server is disabled.
If two or more directory paths overlap between
different servers an error is raised.
"""
# Populate servers
if self._servers is not None:
return
self._servers = {}
# Cycle all the available configurations sections
for section in self._config.sections():
if section == 'barman':
# skip global settings
continue
# Exit if the section has a reserved name
if section in FORBIDDEN_SERVER_NAMES:
msg = "the reserved word '%s' is not allowed as server name." \
"Please rename it." % section
_logger.fatal(msg)
raise SystemExit("FATAL: %s" % msg)
# Create a ServerConfig object
self._servers[section] = ServerConfig(self, section)
# Check for conflicting paths in Barman configuration
self._check_conflicting_paths()
def _check_conflicting_paths(self):
"""
Look for conflicting paths intra-server and inter-server
"""
# All paths in configuration
servers_paths = {}
# Global errors list
self.servers_msg_list = []
# Cycle all the available configurations sections
for section in sorted(self._config.sections()):
if section == 'barman':
# skip global settings
continue
# Paths map
section_conf = self._servers[section]
config_paths = {
'backup_directory':
section_conf.backup_directory,
'basebackups_directory':
section_conf.basebackups_directory,
'errors_directory':
section_conf.errors_directory,
'incoming_wals_directory':
section_conf.incoming_wals_directory,
'streaming_wals_directory':
section_conf.streaming_wals_directory,
'wals_directory':
section_conf.wals_directory,
}
# Check for path errors
for label, path in sorted(config_paths.items()):
# If the path does not conflict with the others, add it to the
# paths map
real_path = os.path.realpath(path)
if real_path not in servers_paths:
servers_paths[real_path] = PathConflict(label, section)
else:
if section == servers_paths[real_path].server:
# Internal path error.
# Insert the error message into the server.msg_list
if real_path == path:
self._servers[section].msg_list.append(
"Conflicting path: %s=%s conflicts with "
"'%s' for server '%s'" % (
label, path,
servers_paths[real_path].label,
servers_paths[real_path].server))
else:
# Symbolic link
self._servers[section].msg_list.append(
"Conflicting path: %s=%s (symlink to: %s) "
"conflicts with '%s' for server '%s'" % (
label, path, real_path,
servers_paths[real_path].label,
servers_paths[real_path].server))
# Disable the server
self._servers[section].disabled = True
else:
# Global path error.
# Insert the error message into the global msg_list
if real_path == path:
self.servers_msg_list.append(
"Conflicting path: "
"%s=%s for server '%s' conflicts with "
"'%s' for server '%s'" % (
label, path, section,
servers_paths[real_path].label,
servers_paths[real_path].server))
else:
# Symbolic link
self.servers_msg_list.append(
"Conflicting path: "
"%s=%s (symlink to: %s) for server '%s' "
"conflicts with '%s' for server '%s'" % (
label, path, real_path, section,
servers_paths[real_path].label,
servers_paths[real_path].server))
def server_names(self):
"""This method returns a list of server names"""
self._populate_servers()
return self._servers.keys()
def servers(self):
"""This method returns a list of server parameters"""
self._populate_servers()
return self._servers.values()
def get_server(self, name):
"""
Get the configuration of the specified server
:param str name: the server name
"""
self._populate_servers()
return self._servers.get(name, None)
def validate_global_config(self):
"""
Validate global configuration parameters
"""
# Check for the existence of unexpected parameters in the
# global section of the configuration file
keys = ['barman_home',
'barman_lock_directory',
'barman_user',
'log_file',
'log_level',
'configuration_files_directory']
keys.extend(ServerConfig.KEYS)
self._validate_with_keys(self._global_config,
keys, 'barman')
def validate_server_config(self, server):
"""
Validate configuration parameters for a specified server
:param str server: the server name
"""
# Check for the existence of unexpected parameters in the
# server section of the configuration file
self._validate_with_keys(self._config.items(server),
ServerConfig.KEYS, server)
@staticmethod
def _validate_with_keys(config_items, allowed_keys, section):
"""
Check every config parameter against a list of allowed keys
:param config_items: list of tuples containing provided parameters
along with their values
:param allowed_keys: list of allowed keys
:param section: source section (for error reporting)
"""
for parameter in config_items:
# if the parameter name is not in the list of allowed values,
# then output a warning
name = parameter[0]
if name not in allowed_keys:
output.warning('Invalid configuration option "%s" in [%s] '
'section.', name, section)
# easy raw config diagnostic with python -m
# noinspection PyProtectedMember
def _main():
print("Active configuration settings:")
r = Config()
r.load_configuration_files_directory()
for section in r._config.sections():
print("Section: %s" % section)
for option in r._config.options(section):
print("\t%s = %s " % (option, r.get(section, option)))
if __name__ == "__main__":
_main()
barman-2.1/barman/infofile.py 0000664 0001750 0001750 00000060071 13015546105 016623 0 ustar gcalacoci gcalacoci # Copyright (C) 2013-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
import ast
import collections
import logging
import os
import dateutil.parser
import dateutil.tz
from barman import xlog
from barman.compression import identify_compression
from barman.exceptions import BackupInfoBadInitialisation
# Named tuple representing a Tablespace with 'name' 'oid' and 'location'
# as property.
Tablespace = collections.namedtuple('Tablespace', 'name oid location')
# Named tuple representing a file 'path' with an associated 'file_type'
TypedFile = collections.namedtuple('ConfFile', 'file_type path')
_logger = logging.getLogger(__name__)
def output_tablespace_list(tablespaces):
"""
Return the literal representation of tablespaces as a Python string
:param tablespaces tablespaces: list of Tablespaces objects
:return str: Literal representation of tablespaces
"""
if tablespaces:
return repr([tuple(item) for item in tablespaces])
else:
return None
def load_tablespace_list(string):
"""
Load the tablespaces as a Python list of namedtuple
Uses ast to evaluate information about tablespaces.
The returned list is used to create a list of namedtuple
:param str string:
:return list: list of namedtuple representing all the tablespaces
"""
obj = ast.literal_eval(string)
if obj:
return [Tablespace._make(item) for item in obj]
else:
return None
def null_repr(obj):
"""
Return the literal representation of an object
:param object obj: object to represent
:return str|None: Literal representation of an object or None
"""
return repr(obj) if obj else None
def load_datetime_tz(time_str):
"""
Load datetime and ensure the result is timezone-aware.
If the parsed timestamp is naive, transform it into a timezone-aware one
using the local timezone.
:param str time_str: string representing a timestamp
:return datetime: the parsed timezone-aware datetime
"""
# dateutil parser returns naive or tz-aware string depending on the format
# of the input string
timestamp = dateutil.parser.parse(time_str)
# if the parsed timestamp is naive, forces it to local timezone
if timestamp.tzinfo is None:
timestamp = timestamp.replace(tzinfo=dateutil.tz.tzlocal())
return timestamp
class Field(object):
def __init__(self, name, dump=None, load=None, default=None, doc=None):
"""
Field descriptor to be used with a FieldListFile subclass.
The resulting field is like a normal attribute with
two optional associated function: to_str and from_str
The Field descriptor can also be used as a decorator
class C(FieldListFile):
x = Field('x')
@x.dump
def x(val): return '0x%x' % val
@x.load
def x(val): return int(val, 16)
:param str name: the name of this attribute
:param callable dump: function used to dump the content to a disk
:param callable load: function used to reload the content from disk
:param default: default value for the field
:param str doc: docstring of the filed
"""
self.name = name
self.to_str = dump
self.from_str = load
self.default = default
self.__doc__ = doc
# noinspection PyUnusedLocal
def __get__(self, obj, objtype=None):
if obj is None:
return self
if not hasattr(obj, '_fields'):
obj._fields = {}
return obj._fields.setdefault(self.name, self.default)
def __set__(self, obj, value):
if not hasattr(obj, '_fields'):
obj._fields = {}
obj._fields[self.name] = value
def __delete__(self, obj):
raise AttributeError("can't delete attribute")
def dump(self, to_str):
return type(self)(self.name, to_str, self.from_str, self.__doc__)
def load(self, from_str):
return type(self)(self.name, self.to_str, from_str, self.__doc__)
class FieldListFile(object):
__slots__ = ('_fields', 'filename')
def __init__(self, **kwargs):
"""
Represent a predefined set of keys with the associated value.
The constructor build the object assigning every keyword argument to
the corresponding attribute. If a provided keyword argument doesn't
has a corresponding attribute an AttributeError exception is raised.
The values provided to the constructor must be of the appropriate
type for the corresponding attribute.
The constructor will not attempt any validation or conversion on them.
This class is meant to be an abstract base class.
:raises: AttributeError
"""
self._fields = {}
self.filename = None
for name in kwargs:
field = getattr(type(self), name, None)
if isinstance(field, Field):
setattr(self, name, kwargs[name])
else:
raise AttributeError('unknown attribute %s' % name)
@classmethod
def from_meta_file(cls, filename):
"""
Factory method that read the specified file and build
an object with its content.
:param str filename: the file to read
"""
o = cls()
o.load(filename)
return o
def save(self, filename=None, file_object=None):
"""
Serialize the object to the specified file or file object
If a file_object is specified it will be used.
If the filename is not specified it uses the one memorized in the
filename attribute. If neither the filename attribute and parameter are
set a ValueError exception is raised.
:param str filename: path of the file to write
:param file file_object: a file like object to write in
:param str filename: the file to write
:raises: ValueError
"""
if file_object:
info = file_object
else:
filename = filename or self.filename
if filename:
info = open(filename, 'w')
else:
info = None
if not info:
raise ValueError(
'either a valid filename or a file_object must be specified')
with info:
for name in sorted(vars(type(self))):
field = getattr(type(self), name)
value = getattr(self, name, None)
if isinstance(field, Field):
if callable(field.to_str):
value = field.to_str(value)
info.write("%s=%s\n" % (name, value))
def load(self, filename=None, file_object=None):
"""
Replaces the current object content with the one deserialized from
the provided file.
This method set the filename attribute.
A ValueError exception is raised if the provided file contains any
invalid line.
:param str filename: path of the file to read
:param file file_object: a file like object to read from
:param str filename: the file to read
:raises: ValueError
"""
if file_object:
info = file_object
elif filename:
info = open(filename, 'r')
else:
raise ValueError(
'either filename or file_object must be specified')
# detect the filename if a file_object is passed
if not filename and file_object:
if hasattr(file_object, 'name'):
filename = file_object.name
# canonicalize filename
if filename:
self.filename = os.path.abspath(filename)
else:
self.filename = None
filename = '' # This is only for error reporting
with info:
for line in info:
# skip spaces and comments
if line.isspace() or line.rstrip().startswith('#'):
continue
# parse the line of form "key = value"
try:
name, value = [x.strip() for x in line.split('=', 1)]
except ValueError:
raise ValueError('invalid line %s in file %s' % (
line.strip(), filename))
# use the from_str function to parse the value
field = getattr(type(self), name, None)
if value == 'None':
value = None
elif isinstance(field, Field) and callable(field.from_str):
value = field.from_str(value)
setattr(self, name, value)
def items(self):
"""
Return a generator returning a list of (key, value) pairs.
If a filed has a dump function defined, it will be used.
"""
for name in sorted(vars(type(self))):
field = getattr(type(self), name)
value = getattr(self, name, None)
if isinstance(field, Field):
if callable(field.to_str):
value = field.to_str(value)
yield (name, value)
def __repr__(self):
return "%s(%s)" % (
self.__class__.__name__,
', '.join(['%s=%r' % x for x in self.items()]))
class WalFileInfo(FieldListFile):
"""
Metadata of a WAL file.
"""
__slots__ = ('orig_filename',)
name = Field('name', doc='base name of WAL file')
size = Field('size', load=int, doc='WAL file size after compression')
time = Field('time', load=float, doc='WAL file modification time '
'(seconds since epoch)')
compression = Field('compression', doc='compression type')
@classmethod
def from_file(cls, filename, default_compression=None, **kwargs):
"""
Factory method to generate a WalFileInfo from a WAL file.
Every keyword argument will override any attribute from the provided
file. If a keyword argument doesn't has a corresponding attribute
an AttributeError exception is raised.
:param str filename: the file to inspect
:param str default_compression: the compression to set if
the current schema is not identifiable.
"""
stat = os.stat(filename)
kwargs.setdefault('name', os.path.basename(filename))
kwargs.setdefault('size', stat.st_size)
kwargs.setdefault('time', stat.st_mtime)
if 'compression' not in kwargs:
kwargs['compression'] = identify_compression(filename) \
or default_compression
obj = cls(**kwargs)
obj.filename = "%s.meta" % filename
obj.orig_filename = filename
return obj
def to_xlogdb_line(self):
"""
Format the content of this object as a xlogdb line.
"""
return "%s\t%s\t%s\t%s\n" % (
self.name,
self.size,
self.time,
self.compression)
@classmethod
def from_xlogdb_line(cls, line):
"""
Parse a line from xlog catalogue
:param str line: a line in the wal database to parse
:rtype: WalFileInfo
"""
try:
name, size, time, compression = line.split()
except ValueError:
# Old format compatibility (no compression)
compression = None
try:
name, size, time = line.split()
except ValueError:
raise ValueError("cannot parse line: %r" % (line,))
# The to_xlogdb_line method writes None values as literal 'None'
if compression == 'None':
compression = None
size = int(size)
time = float(time)
return cls(name=name, size=size, time=time,
compression=compression)
def to_json(self):
"""
Return an equivalent dictionary that can be encoded in json
"""
return dict(self.items())
def relpath(self):
"""
Returns the WAL file path relative to the server's wals_directory
"""
return os.path.join(xlog.hash_dir(self.name), self.name)
def fullpath(self, server):
"""
Returns the WAL file full path
:param barman.server.Server server: the server that owns the wal file
"""
return os.path.join(server.config.wals_directory, self.relpath())
class BackupInfo(FieldListFile):
#: Conversion to string
EMPTY = 'EMPTY'
STARTED = 'STARTED'
FAILED = 'FAILED'
DONE = 'DONE'
STATUS_ALL = (EMPTY, STARTED, DONE, FAILED)
STATUS_NOT_EMPTY = (STARTED, DONE, FAILED)
STATUS_ARCHIVING = (STARTED, DONE)
#: Status according to retention policies
OBSOLETE = 'OBSOLETE'
VALID = 'VALID'
POTENTIALLY_OBSOLETE = 'OBSOLETE*'
NONE = '-'
RETENTION_STATUS = (OBSOLETE, VALID, POTENTIALLY_OBSOLETE, NONE)
version = Field('version', load=int)
pgdata = Field('pgdata')
# Parse the tablespaces as a literal Python list of namedtuple
# Output the tablespaces as a literal Python list of tuple
tablespaces = Field('tablespaces', load=load_tablespace_list,
dump=output_tablespace_list)
# Timeline is an integer
timeline = Field('timeline', load=int)
begin_time = Field('begin_time', load=load_datetime_tz)
begin_xlog = Field('begin_xlog')
begin_wal = Field('begin_wal')
begin_offset = Field('begin_offset', load=int)
size = Field('size', load=int)
deduplicated_size = Field('deduplicated_size', load=int)
end_time = Field('end_time', load=load_datetime_tz)
end_xlog = Field('end_xlog')
end_wal = Field('end_wal')
end_offset = Field('end_offset', load=int)
status = Field('status', default=EMPTY)
server_name = Field('server_name')
error = Field('error')
mode = Field('mode')
config_file = Field('config_file')
hba_file = Field('hba_file')
ident_file = Field('ident_file')
included_files = Field('included_files',
load=ast.literal_eval, dump=null_repr)
backup_label = Field('backup_label', load=ast.literal_eval, dump=null_repr)
__slots__ = ('server', 'config', 'backup_manager',
'backup_id', 'backup_version')
def __init__(self, server, info_file=None, backup_id=None, **kwargs):
"""
Stores meta information about a single backup
:param Server server:
:param file,str,None info_file:
:param str,None backup_id:
:raise BackupInfoBadInitialisation: if the info_file content is invalid
or neither backup_info or
"""
# Initialises the attributes for the object
# based on the predefined keys
super(BackupInfo, self).__init__(**kwargs)
self.server = server
self.config = server.config
self.backup_manager = self.server.backup_manager
self.server_name = self.config.name
self.mode = self.backup_manager.mode
if backup_id:
# Cannot pass both info_file and backup_id
if info_file:
raise BackupInfoBadInitialisation(
'both info_file and backup_id parameters are set')
self.backup_id = backup_id
self.filename = self.get_filename()
# Check if a backup info file for a given server and a given ID
# already exists. If so load the values from the file.
if os.path.exists(self.filename):
self.load(filename=self.filename)
elif info_file:
if hasattr(info_file, 'read'):
# We have been given a file-like object
self.load(file_object=info_file)
else:
# Just a file name
self.load(filename=info_file)
self.backup_id = self.detect_backup_id()
elif not info_file:
raise BackupInfoBadInitialisation(
'backup_id and info_file parameters are both unset')
# Manage backup version for new backup structure
self.backup_version = 2
try:
# the presence of pgdata directory is the marker of version 1
if self.backup_id is not None and os.path.exists(
os.path.join(self.get_basebackup_directory(), 'pgdata')):
self.backup_version = 1
except Exception as e:
_logger.warning("Error detecting backup_version, "
"use default: 2. Failure reason: %s", e)
def get_required_wal_segments(self):
"""
Get the list of required WAL segments for the current backup
"""
return xlog.generate_segment_names(self.begin_wal, self.end_wal,
self.version)
def get_list_of_files(self, target):
"""
Get the list of files for the current backup
"""
# Walk down the base backup directory
if target in ('data', 'standalone', 'full'):
for root, _, files in os.walk(self.get_basebackup_directory()):
for f in files:
yield os.path.join(root, f)
if target in 'standalone':
# List all the WAL files for this backup
for x in self.get_required_wal_segments():
yield self.server.get_wal_full_path(x)
if target in ('wal', 'full'):
for wal_info in self.server.get_wal_until_next_backup(
self,
include_history=True):
yield wal_info.fullpath(self.server)
def detect_backup_id(self):
"""
Detect the backup ID from the name of the parent dir of the info file
"""
if self.filename:
return os.path.basename(os.path.dirname(self.filename))
else:
return None
def get_basebackup_directory(self):
"""
Get the default filename for the backup.info file based on
backup ID and server directory for base backups
"""
return os.path.join(self.config.basebackups_directory,
self.backup_id)
def get_data_directory(self, tablespace_oid=None):
"""
Get path to the backup data dir according with the backup version
If tablespace_oid is passed, build the path to the tablespace
base directory, according with the backup version
:param int tablespace_oid: the oid of a valid tablespace
"""
# Check if a tablespace oid is passed and if is a valid oid
if tablespace_oid is not None and (
self.tablespaces is None or
all(str(tablespace_oid) != str(tablespace.oid)
for tablespace in self.tablespaces)):
raise ValueError("Invalid tablespace OID %s" % tablespace_oid)
# Build the requested path according to backup_version value
path = [self.get_basebackup_directory()]
# Check te version of the backup
if self.backup_version == 2:
# If an oid has been provided, we are looking for a tablespace
if tablespace_oid is not None:
# Append the oid to the basedir of the backup
path.append(str(tablespace_oid))
else:
# Looking for the data dir
path.append('data')
else:
# Backup v1, use pgdata as base
path.append('pgdata')
# If a oid has been provided, we are looking for a tablespace.
if tablespace_oid is not None:
# Append the path to pg_tblspc/oid folder inside pgdata
path.extend(('pg_tblspc', str(tablespace_oid)))
# Return the built path
return os.path.join(*path)
def get_external_config_files(self):
"""
Identify all the configuration files that reside outside the PGDATA.
Returns a list of TypedFile objects.
:rtype: list[TypedFile]
"""
config_files = []
for file_type in ('config_file', 'hba_file', 'ident_file'):
config_file = getattr(self, file_type, None)
if config_file:
# Consider only those that reside outside of the original
# PGDATA directory
if config_file.startswith(self.pgdata):
_logger.debug("Config file '%s' already in PGDATA",
config_file[len(self.pgdata) + 1:])
continue
config_files.append(TypedFile(file_type, config_file))
# Check for any include directives in PostgreSQL configuration
# Currently, include directives are not supported for files that
# reside outside PGDATA. These files must be manually backed up.
# Barman will emit a warning and list those files
if self.included_files:
for included_file in self.included_files:
if not included_file.startswith(self.pgdata):
config_files.append(TypedFile('include', included_file))
return config_files
def get_filename(self):
"""
Get the default filename for the backup.info file based on
backup ID and server directory for base backups
"""
return os.path.join(self.get_basebackup_directory(), 'backup.info')
def set_attribute(self, key, value):
"""
Set a value for a given key
"""
setattr(self, key, value)
def save(self, filename=None, file_object=None):
if not file_object:
# Make sure the containing directory exists
filename = filename or self.filename
dir_name = os.path.dirname(filename)
if not os.path.exists(dir_name):
os.makedirs(dir_name)
super(BackupInfo, self).save(filename=filename,
file_object=file_object)
def to_dict(self):
"""
Return the backup_info content as a simple dictionary
:return dict:
"""
result = dict(self.items())
result.update(backup_id=self.backup_id, server_name=self.server_name,
mode=self.mode, tablespaces=self.tablespaces)
return result
def to_json(self):
"""
Return an equivalent dictionary that uses only json-supported types
"""
data = self.to_dict()
# Convert fields which need special types not supported by json
if data.get('tablespaces') is not None:
data['tablespaces'] = [list(item)
for item in data['tablespaces']]
if data.get('begin_time') is not None:
data['begin_time'] = data['begin_time'].ctime()
if data.get('end_time') is not None:
data['end_time'] = data['end_time'].ctime()
return data
@classmethod
def from_json(cls, server, json_backup_info):
"""
Factory method that builds a BackupInfo object
from a json dictionary
:param barman.Server server: the server related to the Backup
:param dict json_backup_info: the data set containing values from json
"""
data = dict(json_backup_info)
# Convert fields which need special types not supported by json
if data.get('tablespaces') is not None:
data['tablespaces'] = [Tablespace._make(item)
for item in data['tablespaces']]
if data.get('begin_time') is not None:
data['begin_time'] = load_datetime_tz(data['begin_time'])
if data.get('end_time') is not None:
data['end_time'] = load_datetime_tz(data['end_time'])
# Instantiate a BackupInfo object using the converted fields
return cls(server, **data)
barman-2.1/barman/__init__.py 0000664 0001750 0001750 00000001541 12726221047 016567 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
The main Barman module
"""
from __future__ import absolute_import
from .version import __version__
__config__ = None
__all__ = ['__version__', '__config__']
barman-2.1/barman/backup_executor.py 0000664 0001750 0001750 00000174450 13026732452 020226 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
Backup Executor module
A Backup Executor is a class responsible for the execution
of a backup. Specific implementations of backups are defined by
classes that derive from BackupExecutor (e.g.: backup with rsync
through Ssh).
A BackupExecutor is invoked by the BackupManager for backup operations.
"""
import logging
import os
import re
import shutil
from abc import ABCMeta, abstractmethod
from functools import partial
import dateutil.parser
from distutils.version import LooseVersion as Version
from barman import output, xlog
from barman.command_wrappers import PgBaseBackup
from barman.config import BackupOptions
from barman.copy_controller import RsyncCopyController
from barman.exceptions import (CommandFailedException, DataTransferFailure,
FsOperationFailed, PostgresConnectionError,
PostgresIsInRecovery, SshCommandException)
from barman.fs import UnixRemoteCommand
from barman.infofile import BackupInfo
from barman.remote_status import RemoteStatusMixin
from barman.utils import mkpath, with_metaclass
_logger = logging.getLogger(__name__)
class BackupExecutor(with_metaclass(ABCMeta, RemoteStatusMixin)):
"""
Abstract base class for any backup executors.
"""
def __init__(self, backup_manager, mode=None):
"""
Base constructor
:param barman.backup.BackupManager backup_manager: the BackupManager
assigned to the executor
"""
super(BackupExecutor, self).__init__()
self.backup_manager = backup_manager
self.server = backup_manager.server
self.config = backup_manager.config
self.strategy = None
self._mode = mode
# Holds the action being executed. Used for error messages.
self.current_action = None
def init(self):
"""
Initialise the internal state of the backup executor
"""
self.current_action = "starting backup"
@property
def mode(self):
"""
Property that defines the mode used for the backup.
If a strategy is present, the returned string is a combination
of the mode of the executor and the mode of the strategy
(eg: rsync-exclusive)
:return str: a string describing the mode used for the backup
"""
strategy_mode = self.strategy.mode
if strategy_mode:
return "%s-%s" % (self._mode, strategy_mode)
else:
return self._mode
@abstractmethod
def backup(self, backup_info):
"""
Perform a backup for the server - invoked by BackupManager.backup()
:param barman.infofile.BackupInfo backup_info: backup information
"""
def check(self, check_strategy):
"""
Perform additional checks - invoked by BackupManager.check()
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
def status(self):
"""
Set additional status info - invoked by BackupManager.status()
"""
def fetch_remote_status(self):
"""
Get additional remote status info - invoked by
BackupManager.get_remote_status()
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
return {}
def _purge_unused_wal_files(self, backup_info):
"""
It the provided backup is the first, purge all WAL files before the
backup start.
:param barman.infofile.BackupInfo backup_info: the backup to check
"""
# Do nothing if the begin_wal is not defined yet
if backup_info.begin_wal is None:
return
# If this is the first backup, purge unused WAL files
previous_backup = self.backup_manager.get_previous_backup(
backup_info.backup_id)
if not previous_backup:
output.info("This is the first backup for server %s",
self.config.name)
removed = self.backup_manager.remove_wal_before_backup(
backup_info)
if removed:
# report the list of the removed WAL files
output.info("WAL segments preceding the current backup "
"have been found:", log=False)
for wal_name in removed:
output.info("\t%s from server %s "
"has been removed",
wal_name, self.config.name)
def _parse_ssh_command(ssh_command):
"""
Parse a user provided ssh command to a single command and
a list of arguments
In case of error, the first member of the result (the command) will be None
:param ssh_command: a ssh command provided by the user
:return tuple[str,list[str]]: the command and a list of options
"""
try:
ssh_options = ssh_command.split()
except AttributeError:
return None, []
ssh_command = ssh_options.pop(0)
ssh_options.extend("-o BatchMode=yes -o StrictHostKeyChecking=no".split())
return ssh_command, ssh_options
class PostgresBackupExecutor(BackupExecutor):
"""
Concrete class for backup via pg_basebackup (plain format).
Relies on pg_basebackup command to copy data files from the PostgreSQL
cluster using replication protocol.
"""
def __init__(self, backup_manager):
"""
Constructor
:param barman.backup.BackupManager backup_manager: the BackupManager
assigned to the executor
"""
super(PostgresBackupExecutor, self).__init__(backup_manager,
'postgres')
self.validate_configuration()
self.strategy = PostgresBackupStrategy(self)
def validate_configuration(self):
"""
Validate the configuration for this backup executor.
If the configuration is not compatible this method will disable the
server.
"""
# Check for the correct backup options
if BackupOptions.EXCLUSIVE_BACKUP in self.config.backup_options:
self.config.backup_options.remove(
BackupOptions.EXCLUSIVE_BACKUP)
output.warning(
"'exclusive_backup' is not a valid backup_option "
"using postgres backup_method. "
"Overriding with 'concurrent_backup'.")
# Apply the default backup strategy
if BackupOptions.CONCURRENT_BACKUP not in \
self.config.backup_options:
self.config.backup_options.add(BackupOptions.CONCURRENT_BACKUP)
output.debug("The default backup strategy for "
"postgres backup_method is: concurrent_backup")
# Forbid tablespace_bandwidth_limit option.
# It works only with rsync based backups.
if self.config.tablespace_bandwidth_limit:
self.server.config.disabled = True
# Report the error in the configuration errors message list
self.server.config.msg_list.append(
'tablespace_bandwidth_limit option is not supported by '
'postgres backup_method')
# Forbid reuse_backup option.
# It works only with rsync based backups.
if self.config.reuse_backup in ('copy', 'link'):
self.server.config.disabled = True
# Report the error in the configuration errors message list
self.server.config.msg_list.append(
'reuse_backup option is not supported by '
'postgres backup_method')
# Forbid network_compression option.
# It works only with rsync based backups.
if self.config.network_compression:
self.server.config.disabled = True
# Report the error in the configuration errors message list
self.server.config.msg_list.append(
'network_compression option is not supported by '
'postgres backup_method')
# bandwidth_limit option is supported by pg_basebackup executable
# starting from Postgres 9.4
if self.server.config.bandwidth_limit:
# This method is invoked too early to have a working streaming
# connection. So we avoid caching the result by directly
# invoking fetch_remote_status() instead of get_remote_status()
remote_status = self.fetch_remote_status()
# If pg_basebackup is present and it doesn't support bwlimit
# disable the server.
if remote_status['pg_basebackup_bwlimit'] is False:
self.server.config.disabled = True
# Report the error in the configuration errors message list
self.server.config.msg_list.append(
"bandwidth_limit option is not supported by "
"pg_basebackup version (current: %s, required: 9.4)" %
remote_status['pg_basebackup_version'])
def backup(self, backup_info):
"""
Perform a backup for the server - invoked by BackupManager.backup()
through the generic interface of a BackupExecutor.
This implementation is responsible for performing a backup through the
streaming protocol.
The connection must be made with a superuser or a user having
REPLICATION permissions (see PostgreSQL documentation, Section 20.2),
and pg_hba.conf must explicitly permit the replication connection.
The server must also be configured with enough max_wal_senders to leave
at least one session available for the backup.
:param barman.infofile.BackupInfo backup_info: backup information
"""
try:
# Set data directory and server version
self.strategy.start_backup(backup_info)
backup_info.save()
if backup_info.begin_wal is not None:
output.info("Backup start at xlog location: %s (%s, %08X)",
backup_info.begin_xlog,
backup_info.begin_wal,
backup_info.begin_offset)
else:
output.info("Backup start at xlog location: %s",
backup_info.begin_xlog)
# Start the copy
self.current_action = "copying files"
output.info("Copying files.")
self.backup_copy(backup_info)
output.info("Copy done.")
self.strategy.stop_backup(backup_info)
# If this is the first backup, purge eventually unused WAL files
self._purge_unused_wal_files(backup_info)
except CommandFailedException as e:
_logger.exception(e)
raise
def check(self, check_strategy):
"""
Perform additional checks for PostgresBackupExecutor
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
remote_status = self.get_remote_status()
# Check for the presence of pg_basebackup
check_strategy.result(
self.config.name, 'pg_basebackup',
remote_status['pg_basebackup_installed'])
# remote_status['pg_basebackup_compatible'] is None if
# pg_basebackup cannot be executed and False if it is
# not compatible.
hint = None
if not remote_status['pg_basebackup_compatible']:
pg_version = 'Unknown'
basebackup_version = 'Unknown'
if self.server.streaming is not None:
pg_version = self.server.streaming.server_txt_version
if remote_status['pg_basebackup_version'] is not None:
basebackup_version = remote_status['pg_basebackup_version']
hint = "PostgreSQL version: %s, pg_basebackup version: %s" % (
pg_version, basebackup_version
)
check_strategy.result(
self.config.name, 'pg_basebackup compatible',
remote_status['pg_basebackup_compatible'], hint=hint)
# Skip further checks if the postgres connection doesn't work.
# We assume that this error condition will be reported by
# another check.
postgres = self.server.postgres
if postgres is None or postgres.server_txt_version is None:
return
# We can't backup a cluster with tablespaces if the tablespace
# mapping option is not available in the installed version
# of pg_basebackup.
pg_version = Version(postgres.server_txt_version)
tablespaces_list = postgres.get_tablespaces()
# pg_basebackup supports the tablespace-mapping option,
# so there are no problems in this case
if remote_status['pg_basebackup_tbls_mapping']:
hint = None
check_result = True
# pg_basebackup doesn't support the tablespace-mapping option
# and the data directory contains tablespaces, we can't correctly
# backup it.
elif tablespaces_list:
check_result = False
if pg_version < '9.3':
hint = "pg_basebackup can't be used with tablespaces " \
"and PostgreSQL older than 9.3"
else:
hint = "pg_basebackup 9.4 or higher is required for " \
"tablespaces support"
# Even if pg_basebackup doesn't support the tablespace-mapping
# option, this location can be correctly backed up as doesn't
# have any tablespaces
else:
check_result = True
if pg_version < '9.3':
hint = "pg_basebackup can be used as long as tablespaces " \
"support is not required"
else:
hint = "pg_basebackup 9.4 or higher is required for " \
"tablespaces support"
check_strategy.result(
self.config.name,
'pg_basebackup supports tablespaces mapping',
check_result,
hint=hint
)
def fetch_remote_status(self):
"""
Gather info from the remote server.
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
"""
remote_status = dict.fromkeys(
('pg_basebackup_compatible',
'pg_basebackup_installed',
'pg_basebackup_tbls_mapping',
'pg_basebackup_path',
'pg_basebackup_bwlimit',
'pg_basebackup_version'),
None)
# Test pg_basebackup existence
version_info = PgBaseBackup.get_version_info(
self.server.path)
if version_info['full_path']:
remote_status["pg_basebackup_installed"] = True
remote_status["pg_basebackup_path"] = version_info['full_path']
remote_status["pg_basebackup_version"] = (
version_info['full_version'])
pgbasebackup_version = version_info['major_version']
else:
remote_status["pg_basebackup_installed"] = False
return remote_status
# Is bandwidth limit supported?
if remote_status['pg_basebackup_version'] is not None \
and remote_status['pg_basebackup_version'] < '9.4':
remote_status['pg_basebackup_bwlimit'] = False
else:
remote_status['pg_basebackup_bwlimit'] = True
# Is the tablespace mapping option supported?
if pgbasebackup_version >= '9.4':
remote_status["pg_basebackup_tbls_mapping"] = True
else:
remote_status["pg_basebackup_tbls_mapping"] = False
# Retrieve the PostgreSQL version
pg_version = None
if self.server.streaming is not None:
pg_version = self.server.streaming.server_major_version
# If any of the two versions is unknown, we can't compare them
if pgbasebackup_version is None or pg_version is None:
# Return here. We are unable to retrieve
# pg_basebackup or PostgreSQL versions
return remote_status
# pg_version is not None so transform into a Version object
# for easier comparison between versions
pg_version = Version(pg_version)
# pg_basebackup 9.2 is compatible only with PostgreSQL 9.2.
if "9.2" == pg_version == pgbasebackup_version:
remote_status["pg_basebackup_compatible"] = True
# other versions are compatible with lesser versions of PostgreSQL
# WARNING: The development versions of `pg_basebackup` are considered
# higher than the stable versions here, but this is not an issue
# because it accepts everything that is less than
# the `pg_basebackup` version(e.g. '9.6' is less than '9.6devel')
elif "9.2" < pg_version <= pgbasebackup_version:
remote_status["pg_basebackup_compatible"] = True
else:
remote_status["pg_basebackup_compatible"] = False
return remote_status
def backup_copy(self, backup_info):
"""
Perform the actual copy of the backup using pg_basebackup.
First, manages tablespaces, then copies the base backup
using the streaming protocol.
In case of failure during the execution of the pg_basebackup command
the method raises a DataTransferFailure, this trigger the retrying
mechanism when necessary.
:param barman.infofile.BackupInfo backup_info: backup information
"""
# Make sure the destination directory exists, ensure the
# right permissions to the destination dir
backup_dest = backup_info.get_data_directory()
dest_dirs = [backup_dest]
# Manage tablespaces, we need to handle them now in order to
# be able to relocate them inside the
# destination directory of the basebackup
tbs_map = {}
if backup_info.tablespaces:
for tablespace in backup_info.tablespaces:
source = tablespace.location
destination = backup_info.get_data_directory(tablespace.oid)
tbs_map[source] = destination
dest_dirs.append(destination)
# Prepare the destination directories for pgdata and tablespaces
self._prepare_backup_destination(dest_dirs)
# Retrieve pg_basebackup version information
remote_status = self.get_remote_status()
# If pg_basebackup supports --max-rate set the bandwidth_limit
bandwidth_limit = None
if remote_status['pg_basebackup_bwlimit']:
bandwidth_limit = self.config.bandwidth_limit
pg_basebackup = PgBaseBackup(
connection=self.server.streaming,
destination=backup_dest,
command=remote_status['pg_basebackup_path'],
version=remote_status['pg_basebackup_version'],
app_name=self.config.streaming_backup_name,
tbs_mapping=tbs_map,
bwlimit=bandwidth_limit,
immediate=self.config.immediate_checkpoint,
path=self.server.path,
retry_times=self.config.basebackup_retry_times,
retry_sleep=self.config.basebackup_retry_sleep,
retry_handler=partial(self._retry_handler, dest_dirs))
# Do the actual copy
try:
pg_basebackup()
except CommandFailedException as e:
msg = "data transfer failure on directory '%s'" % \
backup_info.get_data_directory()
raise DataTransferFailure.from_command_error(
'pg_basebackup', e, msg)
def _retry_handler(self, dest_dirs, command, args, kwargs,
attempt, exc):
"""
Handler invoked during a backup in case of retry.
The method simply warn the user of the failure and
remove the already existing directories of the backup.
:param list[str] dest_dirs: destination directories
:param RsyncPgData command: Command object being executed
:param list args: command args
:param dict kwargs: command kwargs
:param int attempt: attempt number (starting from 0)
:param CommandFailedException exc: the exception which caused the
failure
"""
output.warning("Failure executing a backup using pg_basebackup "
"(attempt %s)", attempt)
output.warning("The files copied so far will be removed and "
"the backup process will restart in %s seconds",
self.config.basebackup_retry_sleep)
# Remove all the destination directories and reinit the backup
self._prepare_backup_destination(dest_dirs)
def _prepare_backup_destination(self, dest_dirs):
"""
Prepare the destination of the backup, including tablespaces.
This method is also responsible for removing a directory if
it already exists and for ensuring the correct permissions for
the created directories
:param list[str] dest_dirs: destination directories
"""
for dest_dir in dest_dirs:
# Remove a dir if exists. Ignore eventual errors
shutil.rmtree(dest_dir, ignore_errors=True)
# create the dir
mkpath(dest_dir)
# Ensure the right permissions to the destination directory
# chmod 0700 octal
os.chmod(dest_dir, 448)
class SshBackupExecutor(with_metaclass(ABCMeta, BackupExecutor)):
"""
Abstract base class for any backup executors based on Ssh
remote connections. This class is also a factory for
exclusive/concurrent backup strategy objects.
Raises a SshCommandException if 'ssh_command' is not set.
"""
def __init__(self, backup_manager, mode):
"""
Constructor of the abstract class for backups via Ssh
:param barman.backup.BackupManager backup_manager: the BackupManager
assigned to the executor
"""
super(SshBackupExecutor, self).__init__(backup_manager, mode)
# Retrieve the ssh command and the options necessary for the
# remote ssh access.
self.ssh_command, self.ssh_options = _parse_ssh_command(
backup_manager.config.ssh_command)
# Requires ssh_command to be set
if not self.ssh_command:
raise SshCommandException(
'Missing or invalid ssh_command in barman configuration '
'for server %s' % backup_manager.config.name)
# Apply the default backup strategy
if (BackupOptions.CONCURRENT_BACKUP not in
self.config.backup_options and
BackupOptions.EXCLUSIVE_BACKUP not in
self.config.backup_options):
self.config.backup_options.add(BackupOptions.EXCLUSIVE_BACKUP)
output.debug("The default backup strategy for "
"any ssh based backup_method is: "
"exclusive_backup")
# Depending on the backup options value, create the proper strategy
if BackupOptions.CONCURRENT_BACKUP in self.config.backup_options:
# Concurrent backup strategy
self.strategy = ConcurrentBackupStrategy(self)
else:
# Exclusive backup strategy
self.strategy = ExclusiveBackupStrategy(self)
def _update_action_from_strategy(self):
"""
Update the executor's current action with the one of the strategy.
This is used during exception handling to let the caller know
where the failure occurred.
"""
action = getattr(self.strategy, 'current_action', None)
if action:
self.current_action = action
@abstractmethod
def backup_copy(self, backup_info):
"""
Performs the actual copy of a backup for the server
:param barman.infofile.BackupInfo backup_info: backup information
"""
def backup(self, backup_info):
"""
Perform a backup for the server - invoked by BackupManager.backup()
through the generic interface of a BackupExecutor. This implementation
is responsible for performing a backup through a remote connection
to the PostgreSQL server via Ssh. The specific set of instructions
depends on both the specific class that derives from SshBackupExecutor
and the selected strategy (e.g. exclusive backup through Rsync).
:param barman.infofile.BackupInfo backup_info: backup information
"""
# Start the backup, all the subsequent code must be wrapped in a
# try except block which finally issues a stop_backup command
try:
self.strategy.start_backup(backup_info)
except BaseException:
self._update_action_from_strategy()
raise
try:
# save any metadata changed by start_backup() call
# This must be inside the try-except, because it could fail
backup_info.save()
if backup_info.begin_wal is not None:
output.info("Backup start at xlog location: %s (%s, %08X)",
backup_info.begin_xlog,
backup_info.begin_wal,
backup_info.begin_offset)
else:
output.info("Backup start at xlog location: %s",
backup_info.begin_xlog)
# If this is the first backup, purge eventually unused WAL files
self._purge_unused_wal_files(backup_info)
# Start the copy
self.current_action = "copying files"
output.info("Copying files.")
self.backup_copy(backup_info)
output.info("Copy done.")
# Try again to purge eventually unused WAL files. At this point
# the begin_wal value is surely known. Doing it twice is safe
# because this function is useful only during the first backup.
self._purge_unused_wal_files(backup_info)
except:
# we do not need to do anything here besides re-raising the
# exception. It will be handled in the external try block.
raise
else:
self.current_action = "issuing stop of the backup"
output.info("Asking PostgreSQL server to finalize the backup.")
finally:
try:
self.strategy.stop_backup(backup_info)
except BaseException:
self._update_action_from_strategy()
raise
def check(self, check_strategy):
"""
Perform additional checks for SshBackupExecutor, including
Ssh connection (executing a 'true' command on the remote server)
and specific checks for the given backup strategy.
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
hint = "PostgreSQL server"
cmd = None
minimal_ssh_output = None
try:
cmd = UnixRemoteCommand(self.ssh_command,
self.ssh_options,
path=self.server.path)
minimal_ssh_output = ''.join(cmd.get_last_output())
except FsOperationFailed as e:
hint = str(e).strip()
# Output the result
check_strategy.result(self.config.name, 'ssh', cmd is not None, hint)
# Check if the communication channel is "clean"
if minimal_ssh_output:
check_strategy.result(
self.config.name,
'ssh output clean',
False,
"the configured ssh_command must not add anything "
"to the remote command output")
# If SSH works but PostgreSQL is not responding
if (cmd is not None and
self.server.get_remote_status()['server_txt_version']
is None):
# Check for 'backup_label' presence
last_backup = self.server.get_backup(
self.server.get_last_backup_id(BackupInfo.STATUS_NOT_EMPTY)
)
# Look for the latest backup in the catalogue
if last_backup:
# Get PGDATA and build path to 'backup_label'
backup_label = os.path.join(last_backup.pgdata,
'backup_label')
# Verify that backup_label exists in the remote PGDATA.
# If so, send an alert. Do not show anything if OK.
exists = cmd.exists(backup_label)
if exists:
hint = "Check that the PostgreSQL server is up " \
"and no 'backup_label' file is in PGDATA."
check_strategy.result(self.config.name,
'backup_label', False,
hint)
try:
# Invoke specific checks for the backup strategy
self.strategy.check(check_strategy)
except BaseException:
self._update_action_from_strategy()
raise
def status(self):
"""
Set additional status info for SshBackupExecutor using remote
commands via Ssh, as well as those defined by the given
backup strategy.
"""
try:
# Invoke the status() method for the given strategy
self.strategy.status()
except BaseException:
self._update_action_from_strategy()
raise
def fetch_remote_status(self):
"""
Get remote information on PostgreSQL using Ssh, such as
last archived WAL file
This method does not raise any exception in case of errors,
but set the missing values to None in the resulting dictionary.
:rtype: dict[str, None|str]
"""
remote_status = {}
# Retrieve the last archived WAL using a Ssh connection on
# the remote server and executing an 'ls' command. Only
# for pre-9.4 versions of PostgreSQL.
try:
if self.server.postgres and \
self.server.postgres.server_version < 90400:
remote_status['last_archived_wal'] = None
if self.server.postgres.get_setting('data_directory') and \
self.server.postgres.get_setting('archive_command'):
cmd = UnixRemoteCommand(self.ssh_command,
self.ssh_options,
path=self.server.path)
archive_dir = os.path.join(
self.server.postgres.get_setting('data_directory'),
'pg_xlog', 'archive_status')
out = str(cmd.list_dir_content('-t', archive_dir))
for line in out.splitlines():
if line.endswith('.done'):
name = line[:-5]
if xlog.is_any_xlog_file(name):
remote_status['last_archived_wal'] = name
break
except (PostgresConnectionError, FsOperationFailed) as e:
_logger.warn("Error retrieving PostgreSQL status: %s", e)
return remote_status
class RsyncBackupExecutor(SshBackupExecutor):
"""
Concrete class for backup via Rsync+Ssh.
It invokes PostgreSQL commands to start and stop the backup, depending
on the defined strategy. Data files are copied using Rsync via Ssh.
It heavily relies on methods defined in the SshBackupExecutor class
from which it derives.
"""
PGDATA_EXCLUDE_LIST = [
# Exclude this to avoid log files copy
'/pg_log/*',
# Exclude this for (PostgreSQL < 10) to avoid WAL files copy
'/pg_xlog/*',
# This have been renamed on PostgreSQL 10
'/pg_wal/*',
# We handle this on a different step of the copy
'/global/pg_control',
]
EXCLUDE_LIST = [
# Files: see excludeFiles const in PostgreSQL source
'pgsql_tmp*',
'postgresql.auto.conf.tmp',
'postmaster.pid',
'postmaster.opts',
'recovery.conf',
# Directories: see excludeDirContents const in PostgreSQL source
'pg_dynshmem/*',
'pg_notify/*',
'pg_replslot/*',
'pg_serial/*',
'pg_stat_tmp/*',
'pg_snapshots/*',
'pg_subtrans/*',
]
def __init__(self, backup_manager):
"""
Constructor
:param barman.backup.BackupManager backup_manager: the BackupManager
assigned to the strategy
"""
super(RsyncBackupExecutor, self).__init__(backup_manager, 'rsync')
def backup_copy(self, backup_info):
"""
Perform the actual copy of the backup using Rsync.
First, it copies one tablespace at a time, then the PGDATA directory,
and finally configuration files (if outside PGDATA).
Bandwidth limitation, according to configuration, is applied in
the process.
This method is the core of base backup copy using Rsync+Ssh.
:param barman.infofile.BackupInfo backup_info: backup information
"""
# Retrieve the previous backup metadata, then calculate safe_horizon
previous_backup = self.backup_manager.get_previous_backup(
backup_info.backup_id)
safe_horizon = None
reuse_backup = None
if previous_backup:
# safe_horizon is a tz-aware timestamp because BackupInfo class
# ensures that property
reuse_backup = self.config.reuse_backup
safe_horizon = previous_backup.begin_time
# Create the copy controller object, specific for rsync,
# which will drive all the copy operations. Items to be
# copied are added before executing the copy() method
controller = RsyncCopyController(
path=self.server.path,
ssh_command=self.ssh_command,
ssh_options=self.ssh_options,
network_compression=self.config.network_compression,
reuse_backup=reuse_backup,
safe_horizon=safe_horizon,
retry_times=self.config.basebackup_retry_times,
retry_sleep=self.config.basebackup_retry_sleep,
)
# List of paths to be excluded by the PGDATA copy
exclude_and_protect = []
# Process every tablespace
if backup_info.tablespaces:
for tablespace in backup_info.tablespaces:
# If the tablespace location is inside the data directory,
# exclude and protect it from being copied twice during
# the data directory copy
if tablespace.location.startswith(backup_info.pgdata):
exclude_and_protect += [
tablespace.location[len(backup_info.pgdata):]]
# Exclude and protect the tablespace from being copied again
# during the data directory copy
exclude_and_protect += ["pg_tblspc/%s" % tablespace.oid]
# Make sure the destination directory exists in order for
# smart copy to detect that no file is present there
tablespace_dest = backup_info.get_data_directory(
tablespace.oid)
mkpath(tablespace_dest)
# Add the tablespace directory to the list of objects
# to be copied by the controller.
# NOTE: Barman should archive only the content of directory
# "PG_" + PG_MAJORVERSION + "_" + CATALOG_VERSION_NO
# but CATALOG_VERSION_NO is not easy to retrieve, so we copy
# "PG_" + PG_MAJORVERSION + "_*"
# It could select some spurious directory if a development or
# a beta version have been used, but it's good enough for a
# production system as it filters out other major versions.
controller.add_directory(
label=tablespace.name,
src=':%s/' % tablespace.location,
dst=tablespace_dest,
exclude=['/*'] + self.EXCLUDE_LIST,
include=['/PG_%s_*' %
self.server.postgres.server_major_version],
bwlimit=self.config.get_bwlimit(tablespace),
reuse=self._reuse_path(previous_backup, tablespace),
item_class=controller.TABLESPACE_CLASS,
)
# Make sure the destination directory exists in order for smart copy
# to detect that no file is present there
backup_dest = backup_info.get_data_directory()
mkpath(backup_dest)
# Add the PGDATA directory to the list of objects to be copied
# by the controller
controller.add_directory(
label='pgdata',
src=':%s/' % backup_info.pgdata,
dst=backup_dest,
exclude=self.PGDATA_EXCLUDE_LIST + self.EXCLUDE_LIST,
exclude_and_protect=exclude_and_protect,
bwlimit=self.config.get_bwlimit(),
reuse=self._reuse_path(previous_backup),
item_class=controller.PGDATA_CLASS,
)
# At last copy pg_control
controller.add_file(
label='pg_control',
src=':%s/global/pg_control' % backup_info.pgdata,
dst='%s/global/pg_control' % (backup_dest,),
item_class=controller.PGCONTROL_CLASS,
)
# Copy configuration files (if not inside PGDATA)
external_config_files = backup_info.get_external_config_files()
included_config_files = []
for config_file in external_config_files:
# Add included files to a list, they will be handled later
if config_file.file_type == 'include':
included_config_files.append(config_file)
continue
# If the ident file is missing, it isn't an error condition
# for PostgreSQL.
# Barman is consistent with this behavior.
optional = False
if config_file.file_type == 'ident_file':
optional = True
# Create the actual copy jobs in the controller
controller.add_file(
label=config_file.file_type,
src=':%s' % config_file.path,
dst=backup_dest,
optional=optional,
item_class=controller.CONFIG_CLASS,
)
# Execute the copy
try:
controller.copy()
# TODO: Improve the exception output
except CommandFailedException as e:
msg = "data transfer failure"
raise DataTransferFailure.from_command_error(
'rsync', e, msg)
# Check for any include directives in PostgreSQL configuration
# Currently, include directives are not supported for files that
# reside outside PGDATA. These files must be manually backed up.
# Barman will emit a warning and list those files
if any(included_config_files):
output.warning(
"The usage of include directives is not supported "
"for files that reside outside PGDATA.\n"
"Please manually backup the following files:\n"
"\t%s\n",
"\n\t".join(icf.path for icf in included_config_files)
)
def _reuse_path(self, previous_backup_info, tablespace=None):
"""
If reuse_backup is 'copy' or 'link', builds the path of the directory
to reuse, otherwise always returns None.
If oid is None, it returns the full path of PGDATA directory of
the previous_backup otherwise it returns the path to the specified
tablespace using it's oid.
:param barman.infofile.BackupInfo previous_backup_info: backup to be
reused
:param barman.infofile.Tablespace tablespace: the tablespace to copy
:returns: a string containing the local path with data to be reused
or None
:rtype: str|None
"""
oid = None
if tablespace:
oid = tablespace.oid
if self.config.reuse_backup in ('copy', 'link') and \
previous_backup_info is not None:
try:
return previous_backup_info.get_data_directory(oid)
except ValueError:
return None
class BackupStrategy(with_metaclass(ABCMeta, object)):
"""
Abstract base class for a strategy to be used by a backup executor.
"""
#: Regex for START WAL LOCATION info
START_TIME_RE = re.compile('^START TIME: (.*)', re.MULTILINE)
#: Regex for START TIME info
WAL_RE = re.compile('^START WAL LOCATION: (.*) \(file (.*)\)',
re.MULTILINE)
def __init__(self, executor, mode=None):
"""
Constructor
:param BackupExecutor executor: the BackupExecutor assigned
to the strategy
"""
self.executor = executor
# Holds the action being executed. Used for error messages.
self.current_action = None
self.mode = mode
def start_backup(self, backup_info):
"""
Issue a start of a backup - invoked by BackupExecutor.backup()
:param barman.infofile.BackupInfo backup_info: backup information
"""
# Retrieve PostgreSQL server metadata
self._pg_get_metadata(backup_info)
# Record that we are about to start the backup
self.current_action = "issuing start backup command"
_logger.debug(self.current_action)
@abstractmethod
def stop_backup(self, backup_info):
"""
Issue a stop of a backup - invoked by BackupExecutor.backup()
:param barman.infofile.BackupInfo backup_info: backup information
"""
@abstractmethod
def check(self, check_strategy):
"""
Perform additional checks - invoked by BackupExecutor.check()
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
# noinspection PyMethodMayBeStatic
def status(self):
"""
Set additional status info - invoked by BackupExecutor.status()
"""
def _pg_get_metadata(self, backup_info):
"""
Load PostgreSQL metadata into the backup_info parameter
:param barman.infofile.BackupInfo backup_info: backup information
"""
server = self.executor.server
# Get the PostgreSQL data directory location
self.current_action = 'detecting data directory'
output.debug(self.current_action)
data_directory = server.postgres.get_setting('data_directory')
backup_info.set_attribute('pgdata', data_directory)
# Set server version
backup_info.set_attribute('version', server.postgres.server_version)
# Set configuration files location
cf = server.postgres.get_configuration_files()
for key in cf:
backup_info.set_attribute(key, cf[key])
# Get tablespaces information
self.current_action = 'detecting tablespaces'
output.debug(self.current_action)
tablespaces = server.postgres.get_tablespaces()
if tablespaces and len(tablespaces) > 0:
backup_info.set_attribute('tablespaces', tablespaces)
for item in tablespaces:
msg = "\t%s, %s, %s" % (item.oid, item.name, item.location)
_logger.info(msg)
@staticmethod
def _backup_info_from_start_location(backup_info, start_info):
"""
Fill a backup info with information from a start_backup
:param barman.infofile.BackupInfo backup_info: object representing a
backup
:param DictCursor start_info: the result of the pg_start_backup
command
"""
backup_info.set_attribute('status', "STARTED")
backup_info.set_attribute('begin_time', start_info['timestamp'])
backup_info.set_attribute('begin_xlog', start_info['location'])
# PostgreSQL 9.6+ directly provides the timeline
if start_info.get('timeline') is not None:
backup_info.set_attribute('timeline', start_info['timeline'])
# Take a copy of stop_info because we are going to update it
start_info = start_info.copy()
start_info.update(xlog.location_to_xlogfile_name_offset(
start_info['location'],
start_info['timeline']))
# If file_name and file_offset are available, use them
if (start_info.get('file_name') is not None and
start_info.get('file_offset') is not None):
backup_info.set_attribute('begin_wal',
start_info['file_name'])
backup_info.set_attribute('begin_offset',
start_info['file_offset'])
# If the timeline is still missing, extract it from the file_name
if backup_info.timeline is None:
backup_info.set_attribute(
'timeline',
int(start_info['file_name'][0:8], 16))
@staticmethod
def _backup_info_from_stop_location(backup_info, stop_info):
"""
Fill a backup info with information from a backup stop location
:param barman.infofile.BackupInfo backup_info: object representing a
backup
:param DictCursor stop_info: location info of stop backup
"""
# If file_name or file_offset are missing build them using the stop
# location and the timeline.
if (stop_info.get('file_name') is None or
stop_info.get('file_offset') is None):
# Take a copy of stop_info because we are going to update it
stop_info = stop_info.copy()
# Get the timeline from the stop_info if available, otherwise
# Use the one from the backup_label
timeline = stop_info.get('timeline')
if timeline is None:
timeline = backup_info.timeline
stop_info.update(xlog.location_to_xlogfile_name_offset(
stop_info['location'],
timeline))
backup_info.set_attribute('end_time', stop_info['timestamp'])
backup_info.set_attribute('end_xlog', stop_info['location'])
backup_info.set_attribute('end_wal', stop_info['file_name'])
backup_info.set_attribute('end_offset', stop_info['file_offset'])
def _backup_info_from_backup_label(self, backup_info):
"""
Fill a backup info with information from the backup_label file
:param barman.infofile.BackupInfo backup_info: object representing a
backup
"""
# If backup_label is present in backup_info use it...
if backup_info.backup_label:
backup_label_data = backup_info.backup_label
# ... otherwise load backup info from backup_label file
else:
backup_label_path = os.path.join(backup_info.get_data_directory(),
'backup_label')
with open(backup_label_path) as backup_label_file:
backup_label_data = backup_label_file.read()
# Parse backup label
wal_info = self.WAL_RE.search(backup_label_data)
start_time = self.START_TIME_RE.search(backup_label_data)
if wal_info is None or start_time is None:
raise ValueError("Failure parsing backup_label for backup %s" %
backup_info.backup_id)
# Set data in backup_info from backup_label
backup_info.set_attribute('timeline', int(wal_info.group(2)[0:8], 16))
backup_info.set_attribute('begin_xlog', wal_info.group(1))
backup_info.set_attribute('begin_wal', wal_info.group(2))
backup_info.set_attribute('begin_offset', xlog.parse_lsn(
wal_info.group(1)) % xlog.XLOG_SEG_SIZE)
backup_info.set_attribute('begin_time', dateutil.parser.parse(
start_time.group(1)))
class PostgresBackupStrategy(BackupStrategy):
"""
Concrete class for postgres backup strategy.
This strategy is for PostgresBackupExecutor only and is responsible for
executing pre e post backup operations during a physical backup executed
using pg_basebackup.
"""
def check(self, check_strategy):
"""
Perform additional checks for the Postgres backup strategy
"""
def start_backup(self, backup_info):
"""
Manage the start of an pg_basebackup backup
The method performs all the preliminary operations required for a
backup executed using pg_basebackup to start, gathering information
from postgres and filling the backup_info.
:param barman.infofile.BackupInfo backup_info: backup information
"""
self.current_action = "initialising postgres backup_method"
super(PostgresBackupStrategy, self).start_backup(backup_info)
postgres = self.executor.server.postgres
current_xlog_info = postgres.current_xlog_info
self._backup_info_from_start_location(backup_info, current_xlog_info)
def stop_backup(self, backup_info):
"""
Manage the stop of an pg_basebackup backup
The method retrieves the information necessary for the
backup.info file reading the backup_label file.
Due of the nature of the pg_basebackup, information that are gathered
during the start of a backup performed using rsync, are retrieved
here
:param barman.infofile.BackupInfo backup_info: backup information
"""
self._backup_info_from_backup_label(backup_info)
# Set data in backup_info from current_xlog_info
self.current_action = "stopping postgres backup_method"
output.info("Finalising the backup.")
# Get the current xlog position
postgres = self.executor.server.postgres
current_xlog_info = postgres.current_xlog_info.copy()
self._backup_info_from_stop_location(backup_info, current_xlog_info)
# Ask PostgreSQL to switch to another XLOG file. This is needed
# to archive the transaction log file containing the backup
# end position, which is required to recover from the backup.
postgres = self.executor.server.postgres
try:
postgres.switch_xlog()
except PostgresIsInRecovery:
# Skip switching XLOG if a standby server
pass
# Check for the presence of configuration files outside the PGDATA
external_config = backup_info.get_external_config_files()
if any(external_config):
output.warning("pg_basebackup does not copy the PostgreSQL "
"configuration files that reside outside PGDATA. "
"Please manually backup the following files:\n"
"\t%s\n",
"\n\t".join(ecf.path for ecf in external_config))
class ExclusiveBackupStrategy(BackupStrategy):
"""
Concrete class for exclusive backup strategy.
This strategy is for SshBackupExecutor only and is responsible for
coordinating Barman with PostgreSQL on standard physical backup
operations (known as 'exclusive' backup), such as invoking
pg_start_backup() and pg_stop_backup() on the master server.
"""
def __init__(self, executor):
"""
Constructor
:param BackupExecutor executor: the BackupExecutor assigned
to the strategy
"""
super(ExclusiveBackupStrategy, self).__init__(executor, 'exclusive')
# Make sure that executor is of type SshBackupExecutor
assert isinstance(executor, SshBackupExecutor)
# Make sure that backup_options does not contain 'concurrent'
assert (BackupOptions.CONCURRENT_BACKUP not in
self.executor.config.backup_options)
def start_backup(self, backup_info):
"""
Manage the start of an exclusive backup
The method performs all the preliminary operations required for an
exclusive physical backup to start, as well as preparing the
information on the backup for Barman.
:param barman.infofile.BackupInfo backup_info: backup information
"""
super(ExclusiveBackupStrategy, self).start_backup(backup_info)
label = "Barman backup %s %s" % (
backup_info.server_name, backup_info.backup_id)
# Issue an exclusive start backup command
_logger.debug("Start of exclusive backup")
postgres = self.executor.server.postgres
start_info = postgres.start_exclusive_backup(label)
self._backup_info_from_start_location(backup_info, start_info)
def stop_backup(self, backup_info):
"""
Manage the stop of an exclusive backup
The method informs the PostgreSQL server that the physical
exclusive backup is finished, as well as preparing the information
returned by PostgreSQL for Barman.
:param barman.infofile.BackupInfo backup_info: backup information
"""
self.current_action = "issuing stop backup command"
_logger.debug("Stop of exclusive backup")
stop_info = self.executor.server.postgres.stop_exclusive_backup()
self._backup_info_from_stop_location(backup_info, stop_info)
def check(self, check_strategy):
"""
Perform additional checks for ExclusiveBackupStrategy
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
# Make sure PostgreSQL is not in recovery (i.e. is a master)
if self.executor.server.postgres:
is_in_recovery = self.executor.server.postgres.is_in_recovery
if not is_in_recovery:
check_strategy.result(
self.executor.config.name, 'not in recovery', True)
else:
check_strategy.result(
self.executor.config.name, 'not in recovery', False,
'cannot perform exclusive backup on a standby')
class ConcurrentBackupStrategy(BackupStrategy):
"""
Concrete class for concurrent backup strategy.
This strategy is for SshBackupExecutor only and is responsible for
coordinating Barman with PostgreSQL on concurrent physical backup
operations through the pgespresso extension.
"""
def __init__(self, executor):
"""
Constructor
:param BackupExecutor executor: the BackupExecutor assigned
to the strategy
"""
super(ConcurrentBackupStrategy, self).__init__(executor, 'concurrent')
# Make sure that executor is of type SshBackupExecutor
assert isinstance(executor, SshBackupExecutor)
# Make sure that backup_options contains 'concurrent'
assert (BackupOptions.CONCURRENT_BACKUP in
self.executor.config.backup_options)
# noinspection PyMethodMayBeStatic
def _write_backup_label(self, backup_info):
"""
Write backup_label file inside PGDATA folder
:param barman.infofile.BackupInfo backup_info: tbackup information
"""
label_file = os.path.join(backup_info.get_data_directory(),
'backup_label')
output.debug("Writing backup label: %s" % label_file)
with open(label_file, 'w') as f:
f.write(backup_info.backup_label)
def _write_tablespace_map(self, backup_info):
"""
Write tablespace_map file inside PGDATA folder
:param barman.infofile.BackupInfo backup_info: backup information
"""
map_file = os.path.join(backup_info.get_data_directory(),
'tablespace_map')
output.debug("Writing tablespace map")
with open(map_file, 'w') as f:
for tbs in backup_info.tablespaces:
# In some cases (i.e. PostgreSQL on windows) a tablespace
# can contain a newline or a line feed. PostgreSQL
# pg_basebackup code does the same.
quoted_location = re.sub(r'([\n\r])', r'\\\1', tbs.location)
f.write('%s %s\n' % (tbs.oid, quoted_location))
def start_backup(self, backup_info):
"""
Start of the backup.
The method performs all the preliminary operations required for a
backup to start.
:param barman.infofile.BackupInfo backup_info: backup information
"""
super(ConcurrentBackupStrategy, self).start_backup(backup_info)
label = "Barman backup %s %s" % (
backup_info.server_name, backup_info.backup_id)
pg_version = self.executor.server.postgres.server_version
if pg_version >= 90600:
# On 9.6+ execute native concurrent start backup
_logger.debug("Start of native concurrent backup")
self._concurrent_start_backup(backup_info, label)
else:
# On older Postgres use pgespresso
_logger.debug("Start of concurrent backup with pgespresso")
self._pgespresso_start_backup(backup_info, label)
def stop_backup(self, backup_info):
"""
Stop backup wrapper
:param barman.infofile.BackupInfo backup_info: backup information
"""
pg_version = self.executor.server.postgres.server_version
if pg_version >= 90600:
# On 9.6+ execute native concurrent stop backup
_logger.debug("Stop of native concurrent backup")
self._concurrent_stop_backup(backup_info)
else:
# On older Postgres use pgespresso
_logger.debug("Stop of concurrent backup with pgespresso")
self._pgespresso_stop_backup(backup_info)
# Write backup_label retrieved from postgres connection
self.current_action = "writing backup label"
self._write_backup_label(backup_info)
# Write the tablespaces map only if at least a tablespace is present
if backup_info.tablespaces:
self.current_action = "writing tablespace map"
self._write_tablespace_map(backup_info)
# Ask PostgreSQL to switch to another XLOG file. This is needed
# to archive the transaction log file containing the backup
# end position, which is required to recover from the backup.
postgres = self.executor.server.postgres
try:
postgres.switch_xlog()
except PostgresIsInRecovery:
# Skip switching XLOG if a standby server
pass
def _pgespresso_start_backup(self, backup_info, label):
"""
Start a concurrent backup using pgespresso
:param barman.infofile.BackupInfo backup_info: backup information
"""
postgres = self.executor.server.postgres
backup_info.set_attribute('status', "STARTED")
start_info = postgres.pgespresso_start_backup(label)
backup_info.set_attribute('backup_label', start_info['backup_label'])
self._backup_info_from_backup_label(backup_info)
def _pgespresso_stop_backup(self, backup_info):
"""
Stop a concurrent backup using pgespresso
:param barman.infofile.BackupInfo backup_info: backup information
"""
postgres = self.executor.server.postgres
stop_info = postgres.pgespresso_stop_backup(backup_info.backup_label)
# Obtain a modifiable copy of stop_info object
stop_info = stop_info.copy()
# We don't know the exact backup stop location,
# so we include the whole segment.
stop_info['location'] = xlog.location_from_xlogfile_name_offset(
stop_info['end_wal'], 0xFFFFFF)
self._backup_info_from_stop_location(backup_info, stop_info)
def check(self, check_strategy):
"""
Perform additional checks for ConcurrentBackupStrategy
:param CheckStrategy check_strategy: the strategy for the management
of the results of the various checks
"""
postgres = self.executor.server.postgres
try:
# We execute this check only if the postgres connection is non None
# and the server version is lower than 9.6. On latest PostgreSQL
# there is a native API for concurrent backups.
if postgres and postgres.server_version < 90600:
if postgres.has_pgespresso:
check_strategy.result(self.executor.config.name,
'pgespresso extension', True)
else:
check_strategy.result(self.executor.config.name,
'pgespresso extension', False,
'required for concurrent backups on '
'PostgreSQL %s' %
postgres.server_major_version)
except PostgresConnectionError:
# Skip the check if the postgres connection doesn't work.
# We assume that this error condition will be reported by
# another check.
pass
def _concurrent_start_backup(self, backup_info, label):
"""
Start a concurrent backup using the PostgreSQL 9.6
concurrent backup api
:param barman.infofile.BackupInfo backup_info: backup information
:param str label: the backup label
"""
postgres = self.executor.server.postgres
start_info = postgres.start_concurrent_backup(label)
self._backup_info_from_start_location(backup_info, start_info)
def _concurrent_stop_backup(self, backup_info):
"""
Stop a concurrent backup using the PostgreSQL 9.6
concurrent backup api
:param barman.infofile.BackupInfo backup_info: backup information
"""
postgres = self.executor.server.postgres
stop_info = postgres.stop_concurrent_backup()
backup_info.set_attribute('backup_label', stop_info['backup_label'])
self._backup_info_from_stop_location(backup_info, stop_info)
barman-2.1/barman/hooks.py 0000664 0001750 0001750 00000022116 13015546105 016151 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module contains the logic to run hook scripts
"""
import logging
import time
from barman import version
from barman.command_wrappers import Command
from barman.exceptions import AbortedRetryHookScript, UnknownBackupIdException
_logger = logging.getLogger(__name__)
class HookScriptRunner(object):
def __init__(self, backup_manager, name, phase=None, error=None,
retry=False, **extra_env):
"""
Execute a hook script managing its environment
"""
self.backup_manager = backup_manager
self.name = name
self.extra_env = extra_env
self.phase = phase
self.error = error
self.retry = retry
self.environment = None
self.exit_status = None
self.exception = None
self.script = None
self.reset()
def reset(self):
"""
Reset the status of the class.
"""
self.environment = dict(self.extra_env)
config_file = self.backup_manager.config.config.config_file
self.environment.update({
'BARMAN_VERSION': version.__version__,
'BARMAN_SERVER': self.backup_manager.config.name,
'BARMAN_CONFIGURATION': config_file,
'BARMAN_HOOK': self.name,
'BARMAN_RETRY': str(1 if self.retry else 0),
})
if self.error:
self.environment['BARMAN_ERROR'] = self.error
if self.phase:
self.environment['BARMAN_PHASE'] = self.phase
script_config_name = "%s_%s" % (self.phase, self.name)
else:
script_config_name = self.name
self.script = getattr(self.backup_manager.config, script_config_name,
None)
self.exit_status = None
self.exception = None
def env_from_backup_info(self, backup_info):
"""
Prepare the environment for executing a script
:param BackupInfo backup_info: the backup metadata
"""
try:
previous_backup = self.backup_manager.get_previous_backup(
backup_info.backup_id)
if previous_backup:
previous_backup_id = previous_backup.backup_id
else:
previous_backup_id = ''
except UnknownBackupIdException:
previous_backup_id = ''
self.environment.update({
'BARMAN_BACKUP_DIR': backup_info.get_basebackup_directory(),
'BARMAN_BACKUP_ID': backup_info.backup_id,
'BARMAN_PREVIOUS_ID': previous_backup_id,
'BARMAN_STATUS': backup_info.status,
'BARMAN_ERROR': backup_info.error or '',
})
def env_from_wal_info(self, wal_info, full_path=None, error=None):
"""
Prepare the environment for executing a script
:param WalFileInfo wal_info: the backup metadata
:param str full_path: override wal_info.fullpath() result
:param str|Exception error: An error message in case of failure
"""
self.environment.update({
'BARMAN_SEGMENT': wal_info.name,
'BARMAN_FILE': str(full_path if full_path is not None else
wal_info.fullpath(self.backup_manager.server)),
'BARMAN_SIZE': str(wal_info.size),
'BARMAN_TIMESTAMP': str(wal_info.time),
'BARMAN_COMPRESSION': wal_info.compression or '',
'BARMAN_ERROR': str(error or '')
})
def run(self):
"""
Run a a hook script if configured.
This method must never throw any exception
"""
# noinspection PyBroadException
try:
if self.script:
_logger.debug("Attempt to run %s: %s", self.name, self.script)
cmd = Command(
self.script,
env_append=self.environment,
path=self.backup_manager.server.path,
shell=True, check=False)
self.exit_status = cmd()
if self.exit_status != 0:
details = "%s returned %d\n" \
"Output details:\n" \
% (self.script, self.exit_status)
details += cmd.out
details += cmd.err
_logger.warning(details)
else:
_logger.debug("%s returned %d",
self.script,
self.exit_status)
return self.exit_status
except Exception as e:
_logger.exception('Exception running %s', self.name)
self.exception = e
return None
class RetryHookScriptRunner(HookScriptRunner):
"""
A 'retry' hook script is a special kind of hook script that Barman
tries to run indefinitely until it either returns a SUCCESS or
ABORT exit code.
Retry hook scripts are executed immediately before (pre) and after (post)
the command execution. Standard hook scripts are executed immediately
before (pre) and after (post) the retry hook scripts.
"""
# Failed attempts before sleeping for NAP_TIME seconds
ATTEMPTS_BEFORE_NAP = 5
# Short break after a failure (in seconds)
BREAK_TIME = 3
# Long break (nap, in seconds) after ATTEMPTS_BEFORE_NAP failures
NAP_TIME = 60
# ABORT (and STOP) exit code
EXIT_ABORT_STOP = 63
# ABORT (and CONTINUE) exit code
EXIT_ABORT_CONTINUE = 62
# SUCCESS exit code
EXIT_SUCCESS = 0
def __init__(self, backup_manager, name, phase=None, error=None,
**extra_env):
super(RetryHookScriptRunner, self).__init__(
backup_manager, name, phase, error, retry=True, **extra_env)
def run(self):
"""
Run a a 'retry' hook script, if required by configuration.
Barman will retry to run the script indefinitely until it returns
a EXIT_SUCCESS, or an EXIT_ABORT_CONTINUE, or an EXIT_ABORT_STOP code.
There are BREAK_TIME seconds of sleep between every try.
Every ATTEMPTS_BEFORE_NAP failures, Barman will sleep
for NAP_TIME seconds.
"""
# If there is no script, exit
if self.script is not None:
# Keep track of the number of attempts
attempts = 1
while True:
# Run the script using the standard hook method (inherited)
super(RetryHookScriptRunner, self).run()
# Run the script until it returns EXIT_ABORT_CONTINUE,
# or an EXIT_ABORT_STOP, or EXIT_SUCCESS
if self.exit_status in (self.EXIT_ABORT_CONTINUE,
self.EXIT_ABORT_STOP,
self.EXIT_SUCCESS):
break
# Check for the number of attempts
if attempts <= self.ATTEMPTS_BEFORE_NAP:
attempts += 1
# Take a short break
_logger.debug("Retry again in %d seconds", self.BREAK_TIME)
time.sleep(self.BREAK_TIME)
else:
# Reset the attempt number and take a longer nap
_logger.debug("Reached %d failures. Take a nap "
"then retry again in %d seconds",
self.ATTEMPTS_BEFORE_NAP,
self.NAP_TIME)
attempts = 1
time.sleep(self.NAP_TIME)
# Outside the loop check for the exit code.
if self.exit_status == self.EXIT_ABORT_CONTINUE:
# Warn the user if the script exited with EXIT_ABORT_CONTINUE
# Notify EXIT_ABORT_CONTINUE exit status because success and
# failures are already managed in the superclass run method
_logger.warning("%s was aborted (got exit status %d, "
"Barman resumes)",
self.script,
self.exit_status)
elif self.exit_status == self.EXIT_ABORT_STOP:
# Log the error and raise AbortedRetryHookScript exception
_logger.error("%s was aborted (got exit status %d, "
"Barman requested to stop)",
self.script,
self.exit_status)
raise AbortedRetryHookScript(self)
return self.exit_status
barman-2.1/barman/version.py 0000664 0001750 0001750 00000001413 13033425212 016503 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
'''
This module contains the current Barman version.
'''
__version__ = '2.1'
barman-2.1/barman/lockfile.py 0000664 0001750 0001750 00000022515 13015546105 016621 0 ustar gcalacoci gcalacoci # Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
"""
This module is the lock manager for Barman
"""
import errno
import fcntl
import os
import re
from barman.exceptions import (LockFileBusy, LockFileParsingError,
LockFilePermissionDenied)
class LockFile(object):
"""
Ensures that there is only one process which is running against a
specified LockFile.
It supports the Context Manager interface, allowing the use in with
statements.
with LockFile('file.lock') as locked:
if not locked:
print "failed"
else:
You can also use exceptions on failures
try:
with LockFile('file.lock', True):
except LockFileBusy, e, file:
print "failed to lock %s" % file
"""
LOCK_PATTERN = None
"""
If defined in a subclass, it must be a compiled regular expression
which matches the lock filename.
It must provide named groups for the constructor parameters which produce
the same lock name. I.e.:
>>> ServerWalReceiveLock('/tmp', 'server-name').filename
'/tmp/.server-name-receive-wal.lock'
>>> ServerWalReceiveLock.LOCK_PATTERN = re.compile(
r'\.(?P.+)-receive-wal\.lock')
>>> m = ServerWalReceiveLock.LOCK_PATTERN.match(
'.server-name-receive-wal.lock')
>>> ServerWalReceiveLock('/tmp', **(m.groupdict())).filename
'/tmp/.server-name-receive-wal.lock'
"""
@classmethod
def build_if_matches(cls, path):
"""
Factory method that creates a lock instance if the path matches
the lock filename created by the actual class
:param path: the full path of a LockFile
:return:
"""
# If LOCK_PATTERN is not defined always return None
if not cls.LOCK_PATTERN:
return None
# Matches the provided path against LOCK_PATTERN
lock_directory = os.path.abspath(os.path.dirname(path))
lock_name = os.path.basename(path)
match = cls.LOCK_PATTERN.match(lock_name)
if match:
# Build the lock object for the provided path
return cls(lock_directory, **(match.groupdict()))
return None
def __init__(self, filename, raise_if_fail=True, wait=False):
self.filename = os.path.abspath(filename)
self.fd = None
self.raise_if_fail = raise_if_fail
self.wait = wait
def acquire(self, raise_if_fail=None, wait=None):
"""
Creates and holds on to the lock file.
When raise_if_fail, a LockFileBusy is raised if
the lock is held by someone else and a LockFilePermissionDenied is
raised when the user executing barman have insufficient rights for
the creation of a LockFile.
Returns True if lock has been successfully acquired, False otherwise.
:param bool raise_if_fail: If True raise an exception on failure
:param bool wait: If True issue a blocking request
:returns bool: whether the lock has been acquired
"""
if self.fd:
return True
fd = None
# method arguments take precedence on class parameters
raise_if_fail = raise_if_fail \
if raise_if_fail is not None else self.raise_if_fail
wait = wait if wait is not None else self.wait
try:
# 384 is 0600 in octal, 'rw-------'
fd = os.open(self.filename, os.O_CREAT | os.O_RDWR, 384)
flags = fcntl.LOCK_EX
if not wait:
flags |= fcntl.LOCK_NB
fcntl.flock(fd, flags)
# Once locked, replace the content of the file
os.lseek(fd, 0, os.SEEK_SET)
os.write(fd, ("%s\n" % os.getpid()).encode('ascii'))
# Truncate the file at the current position
os.ftruncate(fd, os.lseek(fd, 0, os.SEEK_CUR))
self.fd = fd
return True
except (OSError, IOError) as e:
if fd:
os.close(fd) # let's not leak file descriptors
if raise_if_fail:
if e.errno in (errno.EAGAIN, errno.EWOULDBLOCK):
raise LockFileBusy(self.filename)
elif e.errno == errno.EACCES:
raise LockFilePermissionDenied(self.filename)
else:
raise
else:
return False
def release(self):
"""
Releases the lock.
If the lock is not held by the current process it does nothing.
"""
if not self.fd:
return
try:
fcntl.flock(self.fd, fcntl.LOCK_UN)
os.close(self.fd)
except (OSError, IOError):
pass
self.fd = None
def __del__(self):
"""
Avoid stale lock files.
"""
self.release()
# Contextmanager interface
def __enter__(self):
return self.acquire()
def __exit__(self, exception_type, value, traceback):
self.release()
def get_owner_pid(self):
"""
Test whether a lock is already held by a process.
Returns the PID of the owner process or None if the lock is available.
:rtype: int|None
:raises LockFileParsingError: when the lock content is garbled
:raises LockFilePermissionDenied: when the lockfile is not accessible
"""
try:
self.acquire(raise_if_fail=True, wait=False)
except LockFileBusy:
try:
# Read the lock content and parse the PID
# NOTE: We cannot read it in the self.acquire method to avoid
# reading the previous locker PID
with open(self.filename, 'r') as file_object:
return int(file_object.readline().strip())
except ValueError as e:
# This should not happen
raise LockFileParsingError(e)
# release the lock and return None
self.release()
return None
class GlobalCronLock(LockFile):
"""
This lock protects cron from multiple executions.
Creates a global '.cron.lock' lock file under the given lock_directory.
"""
def __init__(self, lock_directory):
super(GlobalCronLock, self).__init__(
os.path.join(lock_directory, '.cron.lock'),
raise_if_fail=True)
class ServerBackupLock(LockFile):
"""
This lock protects a server from multiple executions of backup command
Creates a '.-backup.lock' lock file under the given lock_directory
for the named SERVER.
"""
def __init__(self, lock_directory, server_name):
super(ServerBackupLock, self).__init__(
os.path.join(lock_directory, '.%s-backup.lock' % server_name),
raise_if_fail=True)
class ServerCronLock(LockFile):
"""
This lock protects a server from multiple executions of cron command
Creates a '.-cron.lock' lock file under the given lock_directory
for the named SERVER.
"""
def __init__(self, lock_directory, server_name):
super(ServerCronLock, self).__init__(
os.path.join(lock_directory, '.%s-cron.lock' % server_name),
raise_if_fail=True, wait=False)
class ServerXLOGDBLock(LockFile):
"""
This lock protects a server's xlogdb access
Creates a '.-xlogdb.lock' lock file under the given lock_directory
for the named SERVER.
"""
def __init__(self, lock_directory, server_name):
super(ServerXLOGDBLock, self).__init__(
os.path.join(lock_directory, '.%s-xlogdb.lock' % server_name),
raise_if_fail=True, wait=True)
class ServerWalArchiveLock(LockFile):
"""
This lock protects a server from multiple executions of wal-archive command
Creates a '.-archive-wal.lock' lock file under
the given lock_directory for the named SERVER.
"""
def __init__(self, lock_directory, server_name):
super(ServerWalArchiveLock, self).__init__(
os.path.join(lock_directory, '.%s-archive-wal.lock' % server_name),
raise_if_fail=True, wait=False)
class ServerWalReceiveLock(LockFile):
"""
This lock protects a server from multiple executions of receive-wal command
Creates a '.-receive-wal.lock' lock file under
the given lock_directory for the named SERVER.
"""
# TODO: Implement on the other LockFile subclasses
LOCK_PATTERN = re.compile(r'\.(?P.+)-receive-wal\.lock')
def __init__(self, lock_directory, server_name):
super(ServerWalReceiveLock, self).__init__(
os.path.join(lock_directory, '.%s-receive-wal.lock' % server_name),
raise_if_fail=True, wait=False)
barman-2.1/README.rst 0000664 0001750 0001750 00000004221 13020304015 014664 0 ustar gcalacoci gcalacoci Barman, Backup and Recovery Manager for PostgreSQL
==================================================
Barman (Backup and Recovery Manager) is an open-source administration
tool for disaster recovery of PostgreSQL servers written in Python. It
allows your organisation to perform remote backups of multiple servers
in business critical environments to reduce risk and help DBAs during
the recovery phase.
Barman is distributed under GNU GPL 3 and maintained by 2ndQuadrant.
For further information, look at the "Web resources" section below.
Source content
--------------
Here you can find a description of files and directory distributed with
Barman:
- AUTHORS : development team of Barman
- NEWS : release notes
- ChangeLog : log of changes
- LICENSE : GNU GPL3 details
- TODO : our wishlist for Barman
- barman : sources in Python
- doc : tutorial and man pages
- rpm : SPEC files for RHEL distributions
- scripts : auxiliary scripts
- tests : unit tests
Web resources
-------------
- Website : http://www.pgbarman.org/
- Download : http://sourceforge.net/projects/pgbarman/files/
- Documentation : http://www.pgbarman.org/documentation/
- Man page, section 1 : http://docs.pgbarman.org/barman.1.html
- Man page, section 5 : http://docs.pgbarman.org/barman.5.html
- Community support : http://www.pgbarman.org/support/
- Professional support : http://www.2ndquadrant.com/
- Client utilities : https://github.com/2ndquadrant-it/barman-cli
- pgespresso extension : https://github.com/2ndquadrant-it/pgespresso
Licence
-------
Copyright (C) 2011-2016 2ndQuadrant Italia Srl
Barman is free software: you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free
Software Foundation, either version 3 of the License, or (at your
option) any later version.
Barman is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
more details.
You should have received a copy of the GNU General Public License along
with Barman. If not, see http://www.gnu.org/licenses/.
barman-2.1/setup.cfg 0000664 0001750 0001750 00000000371 13033432341 015030 0 ustar gcalacoci gcalacoci [bdist_wheel]
universal = 1
[egg_info]
tag_build =
tag_date = 0
tag_svn_revision = 0
[isort]
known_third_party = setuptools,distutils,argh,argcomplete,dateutil,psycopg2,mock,pytest
known_first_party = barman
skip = .tox
[aliases]
test = pytest
barman-2.1/LICENSE 0000644 0001750 0001750 00000104513 12402340601 014211 0 ustar gcalacoci gcalacoci GNU GENERAL PUBLIC LICENSE
Version 3, 29 June 2007
Copyright (C) 2007 Free Software Foundation, Inc.
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
Preamble
The GNU General Public License is a free, copyleft license for
software and other kinds of works.
The licenses for most software and other practical works are designed
to take away your freedom to share and change the works. By contrast,
the GNU General Public License is intended to guarantee your freedom to
share and change all versions of a program--to make sure it remains free
software for all its users. We, the Free Software Foundation, use the
GNU General Public License for most of our software; it applies also to
any other work released this way by its authors. You can apply it to
your programs, too.
When we speak of free software, we are referring to freedom, not
price. Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
them if you wish), that you receive source code or can get it if you
want it, that you can change the software or use pieces of it in new
free programs, and that you know you can do these things.
To protect your rights, we need to prevent others from denying you
these rights or asking you to surrender the rights. Therefore, you have
certain responsibilities if you distribute copies of the software, or if
you modify it: responsibilities to respect the freedom of others.
For example, if you distribute copies of such a program, whether
gratis or for a fee, you must pass on to the recipients the same
freedoms that you received. You must make sure that they, too, receive
or can get the source code. And you must show them these terms so they
know their rights.
Developers that use the GNU GPL protect your rights with two steps:
(1) assert copyright on the software, and (2) offer you this License
giving you legal permission to copy, distribute and/or modify it.
For the developers' and authors' protection, the GPL clearly explains
that there is no warranty for this free software. For both users' and
authors' sake, the GPL requires that modified versions be marked as
changed, so that their problems will not be attributed erroneously to
authors of previous versions.
Some devices are designed to deny users access to install or run
modified versions of the software inside them, although the manufacturer
can do so. This is fundamentally incompatible with the aim of
protecting users' freedom to change the software. The systematic
pattern of such abuse occurs in the area of products for individuals to
use, which is precisely where it is most unacceptable. Therefore, we
have designed this version of the GPL to prohibit the practice for those
products. If such problems arise substantially in other domains, we
stand ready to extend this provision to those domains in future versions
of the GPL, as needed to protect the freedom of users.
Finally, every program is threatened constantly by software patents.
States should not allow patents to restrict development and use of
software on general-purpose computers, but in those that do, we wish to
avoid the special danger that patents applied to a free program could
make it effectively proprietary. To prevent this, the GPL assures that
patents cannot be used to render the program non-free.
The precise terms and conditions for copying, distribution and
modification follow.
TERMS AND CONDITIONS
0. Definitions.
"This License" refers to version 3 of the GNU General Public License.
"Copyright" also means copyright-like laws that apply to other kinds of
works, such as semiconductor masks.
"The Program" refers to any copyrightable work licensed under this
License. Each licensee is addressed as "you". "Licensees" and
"recipients" may be individuals or organizations.
To "modify" a work means to copy from or adapt all or part of the work
in a fashion requiring copyright permission, other than the making of an
exact copy. The resulting work is called a "modified version" of the
earlier work or a work "based on" the earlier work.
A "covered work" means either the unmodified Program or a work based
on the Program.
To "propagate" a work means to do anything with it that, without
permission, would make you directly or secondarily liable for
infringement under applicable copyright law, except executing it on a
computer or modifying a private copy. Propagation includes copying,
distribution (with or without modification), making available to the
public, and in some countries other activities as well.
To "convey" a work means any kind of propagation that enables other
parties to make or receive copies. Mere interaction with a user through
a computer network, with no transfer of a copy, is not conveying.
An interactive user interface displays "Appropriate Legal Notices"
to the extent that it includes a convenient and prominently visible
feature that (1) displays an appropriate copyright notice, and (2)
tells the user that there is no warranty for the work (except to the
extent that warranties are provided), that licensees may convey the
work under this License, and how to view a copy of this License. If
the interface presents a list of user commands or options, such as a
menu, a prominent item in the list meets this criterion.
1. Source Code.
The "source code" for a work means the preferred form of the work
for making modifications to it. "Object code" means any non-source
form of a work.
A "Standard Interface" means an interface that either is an official
standard defined by a recognized standards body, or, in the case of
interfaces specified for a particular programming language, one that
is widely used among developers working in that language.
The "System Libraries" of an executable work include anything, other
than the work as a whole, that (a) is included in the normal form of
packaging a Major Component, but which is not part of that Major
Component, and (b) serves only to enable use of the work with that
Major Component, or to implement a Standard Interface for which an
implementation is available to the public in source code form. A
"Major Component", in this context, means a major essential component
(kernel, window system, and so on) of the specific operating system
(if any) on which the executable work runs, or a compiler used to
produce the work, or an object code interpreter used to run it.
The "Corresponding Source" for a work in object code form means all
the source code needed to generate, install, and (for an executable
work) run the object code and to modify the work, including scripts to
control those activities. However, it does not include the work's
System Libraries, or general-purpose tools or generally available free
programs which are used unmodified in performing those activities but
which are not part of the work. For example, Corresponding Source
includes interface definition files associated with source files for
the work, and the source code for shared libraries and dynamically
linked subprograms that the work is specifically designed to require,
such as by intimate data communication or control flow between those
subprograms and other parts of the work.
The Corresponding Source need not include anything that users
can regenerate automatically from other parts of the Corresponding
Source.
The Corresponding Source for a work in source code form is that
same work.
2. Basic Permissions.
All rights granted under this License are granted for the term of
copyright on the Program, and are irrevocable provided the stated
conditions are met. This License explicitly affirms your unlimited
permission to run the unmodified Program. The output from running a
covered work is covered by this License only if the output, given its
content, constitutes a covered work. This License acknowledges your
rights of fair use or other equivalent, as provided by copyright law.
You may make, run and propagate covered works that you do not
convey, without conditions so long as your license otherwise remains
in force. You may convey covered works to others for the sole purpose
of having them make modifications exclusively for you, or provide you
with facilities for running those works, provided that you comply with
the terms of this License in conveying all material for which you do
not control copyright. Those thus making or running the covered works
for you must do so exclusively on your behalf, under your direction
and control, on terms that prohibit them from making any copies of
your copyrighted material outside their relationship with you.
Conveying under any other circumstances is permitted solely under
the conditions stated below. Sublicensing is not allowed; section 10
makes it unnecessary.
3. Protecting Users' Legal Rights From Anti-Circumvention Law.
No covered work shall be deemed part of an effective technological
measure under any applicable law fulfilling obligations under article
11 of the WIPO copyright treaty adopted on 20 December 1996, or
similar laws prohibiting or restricting circumvention of such
measures.
When you convey a covered work, you waive any legal power to forbid
circumvention of technological measures to the extent such circumvention
is effected by exercising rights under this License with respect to
the covered work, and you disclaim any intention to limit operation or
modification of the work as a means of enforcing, against the work's
users, your or third parties' legal rights to forbid circumvention of
technological measures.
4. Conveying Verbatim Copies.
You may convey verbatim copies of the Program's source code as you
receive it, in any medium, provided that you conspicuously and
appropriately publish on each copy an appropriate copyright notice;
keep intact all notices stating that this License and any
non-permissive terms added in accord with section 7 apply to the code;
keep intact all notices of the absence of any warranty; and give all
recipients a copy of this License along with the Program.
You may charge any price or no price for each copy that you convey,
and you may offer support or warranty protection for a fee.
5. Conveying Modified Source Versions.
You may convey a work based on the Program, or the modifications to
produce it from the Program, in the form of source code under the
terms of section 4, provided that you also meet all of these conditions:
a) The work must carry prominent notices stating that you modified
it, and giving a relevant date.
b) The work must carry prominent notices stating that it is
released under this License and any conditions added under section
7. This requirement modifies the requirement in section 4 to
"keep intact all notices".
c) You must license the entire work, as a whole, under this
License to anyone who comes into possession of a copy. This
License will therefore apply, along with any applicable section 7
additional terms, to the whole of the work, and all its parts,
regardless of how they are packaged. This License gives no
permission to license the work in any other way, but it does not
invalidate such permission if you have separately received it.
d) If the work has interactive user interfaces, each must display
Appropriate Legal Notices; however, if the Program has interactive
interfaces that do not display Appropriate Legal Notices, your
work need not make them do so.
A compilation of a covered work with other separate and independent
works, which are not by their nature extensions of the covered work,
and which are not combined with it such as to form a larger program,
in or on a volume of a storage or distribution medium, is called an
"aggregate" if the compilation and its resulting copyright are not
used to limit the access or legal rights of the compilation's users
beyond what the individual works permit. Inclusion of a covered work
in an aggregate does not cause this License to apply to the other
parts of the aggregate.
6. Conveying Non-Source Forms.
You may convey a covered work in object code form under the terms
of sections 4 and 5, provided that you also convey the
machine-readable Corresponding Source under the terms of this License,
in one of these ways:
a) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by the
Corresponding Source fixed on a durable physical medium
customarily used for software interchange.
b) Convey the object code in, or embodied in, a physical product
(including a physical distribution medium), accompanied by a
written offer, valid for at least three years and valid for as
long as you offer spare parts or customer support for that product
model, to give anyone who possesses the object code either (1) a
copy of the Corresponding Source for all the software in the
product that is covered by this License, on a durable physical
medium customarily used for software interchange, for a price no
more than your reasonable cost of physically performing this
conveying of source, or (2) access to copy the
Corresponding Source from a network server at no charge.
c) Convey individual copies of the object code with a copy of the
written offer to provide the Corresponding Source. This
alternative is allowed only occasionally and noncommercially, and
only if you received the object code with such an offer, in accord
with subsection 6b.
d) Convey the object code by offering access from a designated
place (gratis or for a charge), and offer equivalent access to the
Corresponding Source in the same way through the same place at no
further charge. You need not require recipients to copy the
Corresponding Source along with the object code. If the place to
copy the object code is a network server, the Corresponding Source
may be on a different server (operated by you or a third party)
that supports equivalent copying facilities, provided you maintain
clear directions next to the object code saying where to find the
Corresponding Source. Regardless of what server hosts the
Corresponding Source, you remain obligated to ensure that it is
available for as long as needed to satisfy these requirements.
e) Convey the object code using peer-to-peer transmission, provided
you inform other peers where the object code and Corresponding
Source of the work are being offered to the general public at no
charge under subsection 6d.
A separable portion of the object code, whose source code is excluded
from the Corresponding Source as a System Library, need not be
included in conveying the object code work.
A "User Product" is either (1) a "consumer product", which means any
tangible personal property which is normally used for personal, family,
or household purposes, or (2) anything designed or sold for incorporation
into a dwelling. In determining whether a product is a consumer product,
doubtful cases shall be resolved in favor of coverage. For a particular
product received by a particular user, "normally used" refers to a
typical or common use of that class of product, regardless of the status
of the particular user or of the way in which the particular user
actually uses, or expects or is expected to use, the product. A product
is a consumer product regardless of whether the product has substantial
commercial, industrial or non-consumer uses, unless such uses represent
the only significant mode of use of the product.
"Installation Information" for a User Product means any methods,
procedures, authorization keys, or other information required to install
and execute modified versions of a covered work in that User Product from
a modified version of its Corresponding Source. The information must
suffice to ensure that the continued functioning of the modified object
code is in no case prevented or interfered with solely because
modification has been made.
If you convey an object code work under this section in, or with, or
specifically for use in, a User Product, and the conveying occurs as
part of a transaction in which the right of possession and use of the
User Product is transferred to the recipient in perpetuity or for a
fixed term (regardless of how the transaction is characterized), the
Corresponding Source conveyed under this section must be accompanied
by the Installation Information. But this requirement does not apply
if neither you nor any third party retains the ability to install
modified object code on the User Product (for example, the work has
been installed in ROM).
The requirement to provide Installation Information does not include a
requirement to continue to provide support service, warranty, or updates
for a work that has been modified or installed by the recipient, or for
the User Product in which it has been modified or installed. Access to a
network may be denied when the modification itself materially and
adversely affects the operation of the network or violates the rules and
protocols for communication across the network.
Corresponding Source conveyed, and Installation Information provided,
in accord with this section must be in a format that is publicly
documented (and with an implementation available to the public in
source code form), and must require no special password or key for
unpacking, reading or copying.
7. Additional Terms.
"Additional permissions" are terms that supplement the terms of this
License by making exceptions from one or more of its conditions.
Additional permissions that are applicable to the entire Program shall
be treated as though they were included in this License, to the extent
that they are valid under applicable law. If additional permissions
apply only to part of the Program, that part may be used separately
under those permissions, but the entire Program remains governed by
this License without regard to the additional permissions.
When you convey a copy of a covered work, you may at your option
remove any additional permissions from that copy, or from any part of
it. (Additional permissions may be written to require their own
removal in certain cases when you modify the work.) You may place
additional permissions on material, added by you to a covered work,
for which you have or can give appropriate copyright permission.
Notwithstanding any other provision of this License, for material you
add to a covered work, you may (if authorized by the copyright holders of
that material) supplement the terms of this License with terms:
a) Disclaiming warranty or limiting liability differently from the
terms of sections 15 and 16 of this License; or
b) Requiring preservation of specified reasonable legal notices or
author attributions in that material or in the Appropriate Legal
Notices displayed by works containing it; or
c) Prohibiting misrepresentation of the origin of that material, or
requiring that modified versions of such material be marked in
reasonable ways as different from the original version; or
d) Limiting the use for publicity purposes of names of licensors or
authors of the material; or
e) Declining to grant rights under trademark law for use of some
trade names, trademarks, or service marks; or
f) Requiring indemnification of licensors and authors of that
material by anyone who conveys the material (or modified versions of
it) with contractual assumptions of liability to the recipient, for
any liability that these contractual assumptions directly impose on
those licensors and authors.
All other non-permissive additional terms are considered "further
restrictions" within the meaning of section 10. If the Program as you
received it, or any part of it, contains a notice stating that it is
governed by this License along with a term that is a further
restriction, you may remove that term. If a license document contains
a further restriction but permits relicensing or conveying under this
License, you may add to a covered work material governed by the terms
of that license document, provided that the further restriction does
not survive such relicensing or conveying.
If you add terms to a covered work in accord with this section, you
must place, in the relevant source files, a statement of the
additional terms that apply to those files, or a notice indicating
where to find the applicable terms.
Additional terms, permissive or non-permissive, may be stated in the
form of a separately written license, or stated as exceptions;
the above requirements apply either way.
8. Termination.
You may not propagate or modify a covered work except as expressly
provided under this License. Any attempt otherwise to propagate or
modify it is void, and will automatically terminate your rights under
this License (including any patent licenses granted under the third
paragraph of section 11).
However, if you cease all violation of this License, then your
license from a particular copyright holder is reinstated (a)
provisionally, unless and until the copyright holder explicitly and
finally terminates your license, and (b) permanently, if the copyright
holder fails to notify you of the violation by some reasonable means
prior to 60 days after the cessation.
Moreover, your license from a particular copyright holder is
reinstated permanently if the copyright holder notifies you of the
violation by some reasonable means, this is the first time you have
received notice of violation of this License (for any work) from that
copyright holder, and you cure the violation prior to 30 days after
your receipt of the notice.
Termination of your rights under this section does not terminate the
licenses of parties who have received copies or rights from you under
this License. If your rights have been terminated and not permanently
reinstated, you do not qualify to receive new licenses for the same
material under section 10.
9. Acceptance Not Required for Having Copies.
You are not required to accept this License in order to receive or
run a copy of the Program. Ancillary propagation of a covered work
occurring solely as a consequence of using peer-to-peer transmission
to receive a copy likewise does not require acceptance. However,
nothing other than this License grants you permission to propagate or
modify any covered work. These actions infringe copyright if you do
not accept this License. Therefore, by modifying or propagating a
covered work, you indicate your acceptance of this License to do so.
10. Automatic Licensing of Downstream Recipients.
Each time you convey a covered work, the recipient automatically
receives a license from the original licensors, to run, modify and
propagate that work, subject to this License. You are not responsible
for enforcing compliance by third parties with this License.
An "entity transaction" is a transaction transferring control of an
organization, or substantially all assets of one, or subdividing an
organization, or merging organizations. If propagation of a covered
work results from an entity transaction, each party to that
transaction who receives a copy of the work also receives whatever
licenses to the work the party's predecessor in interest had or could
give under the previous paragraph, plus a right to possession of the
Corresponding Source of the work from the predecessor in interest, if
the predecessor has it or can get it with reasonable efforts.
You may not impose any further restrictions on the exercise of the
rights granted or affirmed under this License. For example, you may
not impose a license fee, royalty, or other charge for exercise of
rights granted under this License, and you may not initiate litigation
(including a cross-claim or counterclaim in a lawsuit) alleging that
any patent claim is infringed by making, using, selling, offering for
sale, or importing the Program or any portion of it.
11. Patents.
A "contributor" is a copyright holder who authorizes use under this
License of the Program or a work on which the Program is based. The
work thus licensed is called the contributor's "contributor version".
A contributor's "essential patent claims" are all patent claims
owned or controlled by the contributor, whether already acquired or
hereafter acquired, that would be infringed by some manner, permitted
by this License, of making, using, or selling its contributor version,
but do not include claims that would be infringed only as a
consequence of further modification of the contributor version. For
purposes of this definition, "control" includes the right to grant
patent sublicenses in a manner consistent with the requirements of
this License.
Each contributor grants you a non-exclusive, worldwide, royalty-free
patent license under the contributor's essential patent claims, to
make, use, sell, offer for sale, import and otherwise run, modify and
propagate the contents of its contributor version.
In the following three paragraphs, a "patent license" is any express
agreement or commitment, however denominated, not to enforce a patent
(such as an express permission to practice a patent or covenant not to
sue for patent infringement). To "grant" such a patent license to a
party means to make such an agreement or commitment not to enforce a
patent against the party.
If you convey a covered work, knowingly relying on a patent license,
and the Corresponding Source of the work is not available for anyone
to copy, free of charge and under the terms of this License, through a
publicly available network server or other readily accessible means,
then you must either (1) cause the Corresponding Source to be so
available, or (2) arrange to deprive yourself of the benefit of the
patent license for this particular work, or (3) arrange, in a manner
consistent with the requirements of this License, to extend the patent
license to downstream recipients. "Knowingly relying" means you have
actual knowledge that, but for the patent license, your conveying the
covered work in a country, or your recipient's use of the covered work
in a country, would infringe one or more identifiable patents in that
country that you have reason to believe are valid.
If, pursuant to or in connection with a single transaction or
arrangement, you convey, or propagate by procuring conveyance of, a
covered work, and grant a patent license to some of the parties
receiving the covered work authorizing them to use, propagate, modify
or convey a specific copy of the covered work, then the patent license
you grant is automatically extended to all recipients of the covered
work and works based on it.
A patent license is "discriminatory" if it does not include within
the scope of its coverage, prohibits the exercise of, or is
conditioned on the non-exercise of one or more of the rights that are
specifically granted under this License. You may not convey a covered
work if you are a party to an arrangement with a third party that is
in the business of distributing software, under which you make payment
to the third party based on the extent of your activity of conveying
the work, and under which the third party grants, to any of the
parties who would receive the covered work from you, a discriminatory
patent license (a) in connection with copies of the covered work
conveyed by you (or copies made from those copies), or (b) primarily
for and in connection with specific products or compilations that
contain the covered work, unless you entered into that arrangement,
or that patent license was granted, prior to 28 March 2007.
Nothing in this License shall be construed as excluding or limiting
any implied license or other defenses to infringement that may
otherwise be available to you under applicable patent law.
12. No Surrender of Others' Freedom.
If conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License. If you cannot convey a
covered work so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you may
not convey it at all. For example, if you agree to terms that obligate you
to collect a royalty for further conveying from those to whom you convey
the Program, the only way you could satisfy both those terms and this
License would be to refrain entirely from conveying the Program.
13. Use with the GNU Affero General Public License.
Notwithstanding any other provision of this License, you have
permission to link or combine any covered work with a work licensed
under version 3 of the GNU Affero General Public License into a single
combined work, and to convey the resulting work. The terms of this
License will continue to apply to the part which is the covered work,
but the special requirements of the GNU Affero General Public License,
section 13, concerning interaction through a network will apply to the
combination as such.
14. Revised Versions of this License.
The Free Software Foundation may publish revised and/or new versions of
the GNU General Public License from time to time. Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.
Each version is given a distinguishing version number. If the
Program specifies that a certain numbered version of the GNU General
Public License "or any later version" applies to it, you have the
option of following the terms and conditions either of that numbered
version or of any later version published by the Free Software
Foundation. If the Program does not specify a version number of the
GNU General Public License, you may choose any version ever published
by the Free Software Foundation.
If the Program specifies that a proxy can decide which future
versions of the GNU General Public License can be used, that proxy's
public statement of acceptance of a version permanently authorizes you
to choose that version for the Program.
Later license versions may give you additional or different
permissions. However, no additional obligations are imposed on any
author or copyright holder as a result of your choosing to follow a
later version.
15. Disclaimer of Warranty.
THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
16. Limitation of Liability.
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES.
17. Interpretation of Sections 15 and 16.
If the disclaimer of warranty and limitation of liability provided
above cannot be given local legal effect according to their terms,
reviewing courts shall apply local law that most closely approximates
an absolute waiver of all civil liability in connection with the
Program, unless a warranty or assumption of liability accompanies a
copy of the Program in return for a fee.
END OF TERMS AND CONDITIONS
How to Apply These Terms to Your New Programs
If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.
To do so, attach the following notices to the program. It is safest
to attach them to the start of each source file to most effectively
state the exclusion of warranty; and each file should have at least
the "copyright" line and a pointer to where the full notice is found.
Copyright (C)
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see .
Also add information on how to contact you by electronic and paper mail.
If the program does terminal interaction, make it output a short
notice like this when it starts in an interactive mode:
Copyright (C)
This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
This is free software, and you are welcome to redistribute it
under certain conditions; type `show c' for details.
The hypothetical commands `show w' and `show c' should show the appropriate
parts of the General Public License. Of course, your program's commands
might be different; for a GUI interface, you would use an "about box".
You should also get your employer (if you work as a programmer) or school,
if any, to sign a "copyright disclaimer" for the program, if necessary.
For more information on this, and how to apply and follow the GNU GPL, see
.
The GNU General Public License does not permit incorporating your program
into proprietary programs. If your program is a subroutine library, you
may consider it more useful to permit linking proprietary applications with
the library. If this is what you want to do, use the GNU Lesser General
Public License instead of this License. But first, please read
.
barman-2.1/setup.py 0000775 0001750 0001750 00000006373 13015546106 014741 0 ustar gcalacoci gcalacoci #!/usr/bin/env python
#
# barman - Backup and Recovery Manager for PostgreSQL
#
# Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see .
"""Backup and Recovery Manager for PostgreSQL
Barman (Backup and Recovery Manager) is an open-source administration
tool for disaster recovery of PostgreSQL servers written in Python.
It allows your organisation to perform remote backups of multiple
servers in business critical environments to reduce risk and help DBAs
during the recovery phase.
Barman is distributed under GNU GPL 3 and maintained by 2ndQuadrant.
"""
import sys
try:
from setuptools import setup
except ImportError:
from distutils.core import setup
if sys.version_info < (2, 6):
raise SystemExit('ERROR: Barman needs at least python 2.6 to work')
# Depend on pytest_runner only when the tests are actually invoked
needs_pytest = set(['pytest', 'test']).intersection(sys.argv)
pytest_runner = ['pytest_runner'] if needs_pytest else []
setup_requires = pytest_runner
install_requires = [
'psycopg2 >= 2.4.2',
'argh >= 0.21.2, <= 0.26.2',
'python-dateutil',
'argcomplete',
]
if sys.version_info < (2, 7):
install_requires += [
'argparse',
]
barman = {}
with open('barman/version.py', 'r') as fversion:
exec(fversion.read(), barman)
setup(
name='barman',
version=barman['__version__'],
author='2ndQuadrant Italia Srl',
author_email='info@2ndquadrant.it',
url='http://www.pgbarman.org/',
packages=['barman', ],
scripts=['bin/barman', ],
data_files=[
('share/man/man1', ['doc/barman.1']),
('share/man/man5', ['doc/barman.5']),
],
license='GPL-3.0',
description=__doc__.split("\n")[0],
long_description="\n".join(__doc__.split("\n")[2:]),
install_requires=install_requires,
platforms=['Linux', 'Mac OS X'],
classifiers=[
'Environment :: Console',
'Development Status :: 5 - Production/Stable',
'Topic :: System :: Archiving :: Backup',
'Topic :: Database',
'Topic :: System :: Recovery Tools',
'Intended Audience :: System Administrators',
'License :: OSI Approved :: GNU General Public License v3 or later '
'(GPLv3+)',
'Programming Language :: Python',
'Programming Language :: Python :: 2.6',
'Programming Language :: Python :: 2.7',
'Programming Language :: Python :: 3.3',
'Programming Language :: Python :: 3.4',
'Programming Language :: Python :: 3.5',
],
setup_requires=setup_requires,
tests_require=[
'mock',
'pytest-catchlog>=1.2.1',
'pytest-timeout',
'pytest',
],
)
barman-2.1/barman.egg-info/ 0000775 0001750 0001750 00000000000 13033432341 016140 5 ustar gcalacoci gcalacoci barman-2.1/barman.egg-info/SOURCES.txt 0000644 0001750 0001750 00000003767 13033432341 020037 0 ustar gcalacoci gcalacoci AUTHORS
ChangeLog
INSTALL
LICENSE
MANIFEST.in
NEWS
README.rst
setup.cfg
setup.py
barman/__init__.py
barman/backup.py
barman/backup_executor.py
barman/cli.py
barman/command_wrappers.py
barman/compression.py
barman/config.py
barman/copy_controller.py
barman/diagnose.py
barman/exceptions.py
barman/fs.py
barman/hooks.py
barman/infofile.py
barman/lockfile.py
barman/output.py
barman/postgres.py
barman/process.py
barman/recovery_executor.py
barman/remote_status.py
barman/retention_policies.py
barman/server.py
barman/utils.py
barman/version.py
barman/wal_archiver.py
barman/xlog.py
barman.egg-info/PKG-INFO
barman.egg-info/SOURCES.txt
barman.egg-info/dependency_links.txt
barman.egg-info/requires.txt
barman.egg-info/top_level.txt
bin/barman
doc/Makefile
doc/barman.1
doc/barman.1.md
doc/barman.5
doc/barman.5.md
doc/barman.conf
doc/barman.d/ssh-server.conf-template
doc/barman.d/streaming-server.conf-template
doc/manual/00-head.en.md
doc/manual/01-intro.en.md
doc/manual/02-before_you_start.en.md
doc/manual/10-design.en.md
doc/manual/15-system_requirements.en.md
doc/manual/16-installation.en.md
doc/manual/17-configuration.en.md
doc/manual/20-server_setup.en.md
doc/manual/21-preliminary_steps.en.md
doc/manual/22-config_file.en.md
doc/manual/23-wal_streaming.en.md
doc/manual/24-wal_archiving.en.md
doc/manual/25-streaming_backup.en.md
doc/manual/26-rsync_backup.en.md
doc/manual/27-windows-support.en.md
doc/manual/41-global-commands.en.md
doc/manual/42-server-commands.en.md
doc/manual/43-backup-commands.en.md
doc/manual/50-feature-details.en.md
doc/manual/65-troubleshooting.en.md
doc/manual/66-about.en.md
doc/manual/70-feature-matrix.en.md
doc/manual/99-references.en.md
rpm/barman.spec
rpm/rhel5/python-dateutil-1.4.1-remove-embedded-timezone-data.patch
rpm/rhel5/python26-argcomplete.spec
rpm/rhel5/python26-argh.spec
rpm/rhel5/python26-dateutil.spec
rpm/rhel5/python26-psycopg2.spec
rpm/rhel5/setup.cfg.patch
rpm/rhel6/python-argcomplete.spec
rpm/rhel6/python-argh.spec
rpm/rhel7/python-argh.spec
scripts/barman.bash_completion barman-2.1/barman.egg-info/top_level.txt 0000644 0001750 0001750 00000000007 13033432341 020665 0 ustar gcalacoci gcalacoci barman
barman-2.1/barman.egg-info/requires.txt 0000644 0001750 0001750 00000000121 13033432341 020530 0 ustar gcalacoci gcalacoci psycopg2 >= 2.4.2
argh >= 0.21.2, <= 0.26.2
python-dateutil
argcomplete
argparse
barman-2.1/barman.egg-info/dependency_links.txt 0000644 0001750 0001750 00000000001 13033432341 022204 0 ustar gcalacoci gcalacoci
barman-2.1/barman.egg-info/PKG-INFO 0000644 0001750 0001750 00000002504 13033432341 017234 0 ustar gcalacoci gcalacoci Metadata-Version: 1.1
Name: barman
Version: 2.1
Summary: Backup and Recovery Manager for PostgreSQL
Home-page: http://www.pgbarman.org/
Author: 2ndQuadrant Italia Srl
Author-email: info@2ndquadrant.it
License: GPL-3.0
Description: Barman (Backup and Recovery Manager) is an open-source administration
tool for disaster recovery of PostgreSQL servers written in Python.
It allows your organisation to perform remote backups of multiple
servers in business critical environments to reduce risk and help DBAs
during the recovery phase.
Barman is distributed under GNU GPL 3 and maintained by 2ndQuadrant.
Platform: Linux
Platform: Mac OS X
Classifier: Environment :: Console
Classifier: Development Status :: 5 - Production/Stable
Classifier: Topic :: System :: Archiving :: Backup
Classifier: Topic :: Database
Classifier: Topic :: System :: Recovery Tools
Classifier: Intended Audience :: System Administrators
Classifier: License :: OSI Approved :: GNU General Public License v3 or later (GPLv3+)
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 2.6
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 3.3
Classifier: Programming Language :: Python :: 3.4
Classifier: Programming Language :: Python :: 3.5
barman-2.1/AUTHORS 0000664 0001750 0001750 00000002660 13015546105 014266 0 ustar gcalacoci gcalacoci Barman Core Team (in alphabetical order):
* Gabriele Bartolini (project leader)
* Jonathan Battiato (QA/testing)
* Stefano Bianucci (developer, intern from University of Florence)
* Giuseppe Broccolo (QA/testing)
* Giulio Calacoci (developer)
* Francesco Canovai (QA/testing)
* Leonardo Cecchi (developer)
* Gianni Ciolli (QA/testing)
* Britt Cole (documentation)
* Marco Nenciarini (lead developer)
* Rubens Souza (QA/testing)
Past contributors:
* Carlo Ascani
Many thanks go to our sponsors (in alphabetical order):
* 4Caast - http://4caast.morfeo-project.org/ (Founding sponsor)
* Adyen - http://www.adyen.com/
* Agile Business Group - http://www.agilebg.com/
* BIJ12 - http://www.bij12.nl/
* CSI Piemonte - http://www.csipiemonte.it/ (Founding sponsor)
* Ecometer - http://www.ecometer.it/
* GestionaleAuto - http://www.gestionaleauto.com/ (Founding sponsor)
* Jobrapido - http://www.jobrapido.com/
* Navionics - http://www.navionics.com/ (Founding sponsor)
* Sovon Vogelonderzoek Nederland - https://www.sovon.nl/
* Subito.it - http://www.subito.it/
* XCon Internet Services - http://www.xcon.it/ (Founding sponsor)
barman-2.1/INSTALL 0000664 0001750 0001750 00000000344 13015546105 014244 0 ustar gcalacoci gcalacoci Barman INSTALL instructions
Copyright (C) 2011-2016 2ndQuadrant Italia Srl
For further information, see the "Installation" section in the
official manual of Barman or the Markdown source file:
doc/manual/16-installation.en.md.
barman-2.1/NEWS 0000664 0001750 0001750 00000053450 13033425212 013713 0 ustar gcalacoci gcalacoci Barman News - History of user-visible changes
Copyright (C) 2011-2016 2ndQuadrant Italia Srl
Version 2.1 - 5 Jan 2017
- Add --archive and --archive-timeout options to switch-xlog command
- Preliminary support for PostgreSQL 10 (#73)
- Minor additions:
- Add last archived WAL info to diagnose output
- Add start time and execution time to the output of delete
command
- Minor bug fixes:
- Return failure for get-wal command on inactive server
- Make streaming_archiver_names and streaming_backup_name options
global (#57)
- Fix rsync failures due to files truncated during transfer (#64)
- Correctly handle compressed history files (#66)
- Avoid de-referencing symlinks in pg_tblspc when preparing
recovery (#55)
- Fix comparison of last archiving failure (#40, #58)
- Avoid failing recovery if postgresql.conf is not writable (#68)
- Fix output of replication-status command (#56)
- Exclude files from backups like pg_basebackup (#65, #72)
- Exclude directories from other Postgres versions while copying
tablespaces (#74)
- Make retry hook script options global
Version 2.0 - 27 Sep 2016
- Support for pg_basebackup and base backups over the PostgreSQL
streaming replication protocol with backup_method=postgres
(PostgreSQL 9.1 or higher required)
- Support for physical replication slots through the slot_name
configuration option as well as the --create-slot and --drop-slot
options for the receive-wal command (PostgreSQL 9.4 or higher
required). When slot_name is specified and streaming_archiver is
enabled, receive-wal transparently integrates with pg_receivexlog,
and check makes sure that slots exist and are actively used
- Support for the new backup API introduced in PostgreSQL 9.6, which
transparently enables concurrent backups and backups from standby
servers using the standard rsync method of backup. Concurrent backup
was only possible for PostgreSQL 9.2 to 9.5 versions through the
pgespresso extension. The new backup API will make pgespresso
redundant in the future
- If properly configured, Barman can function as a synchronous standby
in terms of WAL streaming. By properly setting the
streaming_archiver_name in the synchronous_standby_names priority
list on the master, and enabling replication slot support, the
receive-wal command can now be part of a PostgreSQL synchronous
replication cluster, bringing RPO=0 (PostgreSQL 9.5.5 or
higher required)
- Introduce barman-wal-restore, a standard and robust script written
in Python that can be used as restore_command in recovery.conf files
of any standby server of a cluster. It supports remote parallel
fetching of WAL files by efficiently invoking get-wal through SSH.
Currently available as a separate project called barman-cli. The
barman-cli package is required for remote recovery when get-wal is
listed in recovery_options
- Control the maximum execution time of the check command through the
check_timeout global/server configuration option (30 seconds
by default)
- Limit the number of WAL segments that are processed by an
archive-wal run, through the archiver_batch_size and
streaming_archiver_batch_size global/server options which control
archiving of WAL segments coming from, respectively, the standard
archiver and receive-wal
- Removed locking of the XLOG database during check operations
- The show-backup command is now aware of timelines and properly
displays which timelines can be used as recovery targets for a given
base backup. Internally, Barman is now capable of parsing .history
files
- Improved the logic behind the retry mechanism when copy operations
experience problems. This involves backup (rsync and postgres) as
well as remote recovery (rsync)
- Code refactoring involving remote command and physical copy
interfaces
- Bug fixes:
- Correctly handle .history files from streaming
- Fix replication-status on PostgreSQL 9.1
- Fix replication-status when sent and write locations are not
available
- Fix misleading message on pg_receivexlog termination
Version 1.6.1 - 23 May 2016
- Add --peek option to get-wal command to discover existing WAL files
from the Barman's archive
- Add replication-status command for monitoring the status of any
streaming replication clients connected to the PostgreSQL server.
The --target option allows users to limit the request to only hot
standby servers or WAL streaming clients
- Add the switch-xlog command to request a switch of a WAL file to the
PostgreSQL server. Through the '--force' it issues a CHECKPOINT
beforehand
- Add streaming_archiver_name option, which sets a proper
application_name to pg_receivexlog when streaming_archiver is
enabled (only for PostgreSQL 9.3 and above)
- Check for _superuser_ privileges with PostgreSQL's standard
connections (#30)
- Check the WAL archive is never empty
- Check for 'backup_label' on the master when server is down
- Improve barman-wal-restore contrib script
- Bug fixes:
- Treat the "failed backups" check as non-fatal
- Rename '-x' option for get-wal as '-z'
- Add archive_mode=always support for PostgreSQL 9.5 (#32)
- Properly close PostgreSQL connections when necessary
- Fix receive-wal for pg_receive_xlog version 9.2
Version 1.6.0 - 29 Feb 2016
- Support for streaming replication connection through the
streaming_conninfo server option
- Support for the streaming_archiver option that allows Barman to
receive WAL files through PostgreSQL's native streaming protocol.
When set to 'on', it relies on pg_receivexlog to receive WAL data,
reducing Recovery Point Objective. Currently, WAL streaming is an
additional feature (standard log archiving is still required)
- Implement the receive-wal command that, when streaming_archiver is
on, wraps pg_receivexlog for WAL streaming. Add --stop option to
stop receiving WAL files via streaming protocol. Add --reset option
to reset the streaming status and restart from the current xlog
in Postgres.
- Automatic management (startup and stop) of receive-wal command via
cron command
- Support for the path_prefix configuration option
- Introduction of the archiver option (currently fixed to on) which
enables continuous WAL archiving for a specific server, through log
shipping via PostgreSQL's archive_command
- Support for streaming_wals_directory and errors_directory options
- Management of WAL duplicates in archive-wal command and integration
with check command
- Verify if pg_receivexlog is running in check command when
streaming_archiver is enabled
- Verify if failed backups are present in check command
- Accept compressed WAL files in incoming directory
- Add support for the pigz compressor (thanks to Stefano Zacchiroli
zack@upsilon.cc)
- Implement pygzip and pybzip2 compressors (based on an initial idea
of Christoph Moench-Tegeder christoph@2ndquadrant.de)
- Creation of an implicit restore point at the end of a backup
- Current size of the PostgreSQL data files in barman status
- Permit archive_mode=always for PostgreSQL 9.5 servers (thanks to
Christoph Moench-Tegeder christoph@2ndquadrant.de)
- Complete refactoring of the code responsible for connecting to
PostgreSQL
- Improve messaging of cron command regarding sub-processes
- Native support for Python >= 3.3
- Changes of behaviour:
- Stop trashing WAL files during archive-wal (commit:e3a1d16)
- Bug fixes:
- Atomic WAL file archiving (#9 and #12)
- Propagate "-c" option to any Barman subprocess (#19)
- Fix management of backup ID during backup deletion (#22)
- Improve archive-wal robustness and log messages (#24)
- Improve error handling in case of missing parameters
Version 1.5.1 - 16 Nov 2015
- Add support for the 'archive-wal' command which performs WAL
maintenance operations on a given server
- Add support for "per-server" concurrency of the 'cron' command
- Improved management of xlog.db errors
- Add support for mixed compression types in WAL files (SF.net#61)
- Bug fixes:
- Avoid retention policy checks during the recovery
- Avoid 'wal_level' check on PostgreSQL version < 9.0 (#3)
- Fix backup size calculation (#5)
Version 1.5.0 - 28 Sep 2015
- Add support for the get-wal command which allows users to fetch any
WAL file from the archive of a specific server
- Add support for retry hook scripts, a special kind of hook scripts
that Barman tries to run until they succeed
- Add active configuration option for a server to temporarily disable
the server by setting it to False
- Add barman_lock_directory global option to change the location of
lock files (by default: 'barman_home')
- Execute the full suite of checks before starting a backup, and skip
it in case one or more checks fail
- Forbid to delete a running backup
- Analyse include directives of a PostgreSQL server during backup and
recover operations
- Add check for conflicting paths in the configuration of Barman, both
intra (by temporarily disabling a server) and inter-server (by
refusing any command, to any server).
- Add check for wal_level
- Add barman-wal-restore script to be used as restore_command on a
standby server, in conjunction with barman get-wal
- Implement a standard and consistent policy for error management
- Improved cache management of backups
- Improved management of configuration in unit tests
- Tutorial and man page sources have been converted to Markdown format
- Add code documentation through Sphinx
- Complete refactor of the code responsible for managing the backup
and the recover commands
- Changed internal directory structure of a backup
- Introduce copy_method option (currently fixed to rsync)
- Bug fixes:
- Manage options without '=' in PostgreSQL configuration files
- Preserve Timeline history files (Fixes: #70)
- Workaround for rsync on SUSE Linux (Closes: #13 and #26)
- Disables dangerous settings in postgresql.auto.conf
(Closes: #68)
Version 1.4.1 - 05 May 2015
* Fix for WAL archival stop working if first backup is EMPTY
(Closes: #64)
* Fix exception during error handling in Barman recovery
(Closes: #65)
* After a backup, limit cron activity to WAL archiving only
(Closes: #62)
* Improved robustness and error reporting of the backup delete
command (Closes: #63)
* Fix computation of WAL production ratio as reported in the
show-backup command
* Improved management of xlogb file, which is now correctly fsynced
when updated. Also, the rebuild-xlogdb command now operates on a
temporary new file, which overwrites the main one when finished.
* Add unit tests for dateutil module compatibility
* Modified Barman version following PEP 440 rules and added support
of tests in Python 3.4
Version 1.4.0 - 26 Jan 2015
* Incremental base backup implementation through the reuse_backup
global/server option. Possible values are off (disabled,
default), copy (preventing unmodified files from being
transferred) and link (allowing for deduplication through hard
links).
* Store and show deduplication effects when using reuse_backup=
link.
* Added transparent support of pg_stat_archiver (PostgreSQL 9.4) in
check, show-server and status commands.
* Improved administration by invoking WAL maintenance at the end of
a successful backup.
* Changed the way unused WAL files are trashed, by differentiating
between concurrent and exclusive backup cases.
* Improved performance of WAL statistics calculation.
* Treat a missing pg_ident.conf as a WARNING rather than an error.
* Refactored output layer by removing remaining yield calls.
* Check that rsync is in the system path.
* Include history files in WAL management.
* Improved robustness through more unit tests.
* Fixed bug #55: Ignore fsync EINVAL errors on directories.
* Fixed bug #58: retention policies delete.
Version 1.3.3 - 21 Aug 2014
* Added "last_backup_max_age", a new global/server option that
allows administrators to set the max age of the last backup in a
catalogue, making it easier to detect any issues with periodical
backup execution
* Improved robustness of "barman backup" by introducing two global/
server options: "basebackup_retry_times" and
"basebackup_retry_sleep". These options allow an administrator to
specify, respectively, the number of attempts for a copy
operation after a failure, and the number of seconds of wait
before retrying
* Improved the recovery process via rsync on an existing directory
(incremental recovery), by splitting the previous rsync call into
several ones - invoking checksum control only when necessary
* Added support for PostgreSQL 8.3
* Minor changes:
+ Support for comma separated list values configuration options
+ Improved backup durability by calling fsync() on backup and
WAL files during "barman backup" and "barman cron"
+ Improved Nagios output for "barman check --nagios"
+ Display compression ratio for WALs in "barman show-backup"
+ Correctly handled keyboard interruption (CTRL-C) while
performing barman backup
+ Improved error messages of failures regarding the stop of a
backup
+ Wider coverage of unit tests
* Bug fixes:
+ Copies "recovery.conf" on the remote server during "barman
recover" (#45)
+ Correctly detect pre/post archive hook scripts (#41)
Version 1.3.2 - 15 Apr 2014
* Fixed incompatibility with PostgreSQL 8.4 (Closes #40, bug
introduced in version 1.3.1)
Version 1.3.1 - 14 Apr 2014
* Added support for concurrent backup of PostgreSQL 9.2 and 9.3
servers that use the "pgespresso" extension. This feature is
controlled by the "backup_options" configuration option (global/
server) and activated when set to "concurrent_backup". Concurrent
backup allows DBAs to perform full backup operations from a
streaming replicated standby.
* Added the "barman diagnose" command which prints important
information about the Barman system (extremely useful for support
and problem solving)
* Improved error messages and exception handling interface
* Fixed bug in recovery of tablespaces that are created inside the
PGDATA directory (bug introduced in version 1.3.0)
* Fixed minor bug of unhandled -q option, for quiet mode of
commands to be used in cron jobs (bug introduced in version
1.3.0)
* Minor bug fixes and code refactoring
Version 1.3.0 - 3 Feb 2014
* Refactored BackupInfo class for backup metadata to use the new
FieldListFile class (infofile module)
* Refactored output layer to use a dedicated module, in order to
facilitate integration with Nagios (NagiosOutputWriter class)
* Refactored subprocess handling in order to isolate stdin/stderr/
stdout channels (command_wrappers module)
* Refactored hook scripts management
* Extracted logging configuration and userid enforcement from the
configuration class.
* Support for hook scripts to be executed before and after a WAL
file is archived, through the 'pre_archive_script' and
'post_archive_script' configuration options.
* Implemented immediate checkpoint capability with
--immediate-checkpoint command option and 'immediate_checkpoint'
configuration option
* Implemented network compression for remote backup and recovery
through the 'network_compression' configuration option (#19)
* Implemented the 'rebuild-xlogdb' command (Closes #27 and #28)
* Added deduplication of tablespaces located inside the PGDATA
directory
* Refactored remote recovery code to work the same way local
recovery does, by performing remote directory preparation
(assuming the remote user has the right permissions on the remote
server)
* 'barman backup' now tries and create server directories before
attempting to execute a full backup (#14)
* Fixed bug #22: improved documentation for tablespaces relocation
* Fixed bug #31: 'barman cron' checks directory permissions for
lock file
* Fixed bug #32: xlog.db read access during cron activities
Version 1.2.3 - 5 September 2013
* Added support for PostgreSQL 9.3
* Added support for the "--target-name" recovery option, which allows to
restore to a named point previously specified with pg_create_restore_point
(only for PostgreSQL 9.1 and above users)
* Fixed bug #27 about flock() usage with barman.lockfile (many thanks to
Damon Snyder )
* Introduced Python 3 compatibility
Version 1.2.2 - 24 June 2013
* Fix python 2.6 compatibility
Version 1.2.1 - 17 June 2013
* Added the "bandwidth_limit" global/server option which allows
to limit the I/O bandwidth (in KBPS) for backup and recovery operations
* Added the "tablespace_bandwidth_limit" global/server option which allows
to limit the I/O bandwidth (in KBPS) for backup and recovery operations
on a per tablespace basis
* Added /etc/barman/barman.conf as default location
* Bug fix: avoid triggering the minimum_redundancy check
on FAILED backups (thanks to JĂ©rĂ´me Vanandruel)
Version 1.2.0 - 31 Jan 2013
* Added the "retention_policy_mode" global/server option which defines
the method for enforcing retention policies (currently only "auto")
* Added the "minimum_redundancy" global/server option which defines
the minimum number of backups to be kept for a server
* Added the "retention_policy" global/server option which defines
retention policies management based on redunancy (e.g. REDUNDANCY 4)
or recovery window (e.g. RECOVERY WINDOW OF 3 MONTHS)
* Added retention policy support to the logging infrastructure, the
"check" and the "status" commands
* The "check" command now integrates minimum redundancy control
* Added retention policy states (valid, obsolete and potentially obsolete)
to "show-backup" and "list-backup" commands
* The 'all' keyword is now forbidden as server name
* Added basic support for Nagios plugin output to the 'check'
command through the --nagios option
* Barman now requires argh => 0.21.2 and argcomplete-
* Minor bug fixes
Version 1.1.2 - 29 Nov 2012
* Added "configuration_files_directory" option that allows
to include multiple server configuration files from a directory
* Support for special backup IDs: latest, last, oldest, first
* Management of multiple servers to the 'list-backup' command.
'barman list-backup all' now list backups for all the configured servers.
* Added "application_name" management for PostgreSQL >= 9.0
* Fixed bug #18: ignore missing WAL files if not found during delete
Version 1.1.1 - 16 Oct 2012
* Fix regressions in recover command.
Version 1.1.0 - 12 Oct 2012
* Support for hook scripts to be executed before and after
a 'backup' command through the 'pre_backup_script' and 'post_backup_script'
configuration options.
* Management of multiple servers to the 'backup' command.
'barman backup all' now iteratively backs up all the configured servers.
* Fixed bug #9: "9.2 issue with pg_tablespace_location()"
* Add warning in recovery when file location options have been defined
in the postgresql.conf file (issue #10)
* Fail fast on recover command if the destination directory contains
the ':' character (Closes: #4) or if an invalid tablespace
relocation rule is passed
* Report an informative message when pg_start_backup() invocation
fails because an exclusive backup is already running (Closes: #8)
Version 1.0.0 - 6 July 2012
* Backup of multiple PostgreSQL servers, with different versions. Versions
from PostgreSQL 8.4+ are supported.
* Support for secure remote backup (through SSH)
* Management of a catalog of backups for every server, allowing users
to easily create new backups, delete old ones or restore them
* Compression of WAL files that can be configured on a per server
basis using compression/decompression filters, both predefined (gzip
and bzip2) or custom
* Support for INI configuration file with global and per-server directives.
Default location for configuration files are /etc/barman.conf or
~/.barman.conf. The '-c' option allows users to specify a different one
* Simple indexing of base backups and WAL segments that does not require
a local database
* Maintenance mode (invoked through the 'cron' command) which performs
ordinary operations such as WAL archival and compression, catalog
updates, etc.
* Added the 'backup' command which takes a full physical base backup
of the given PostgreSQL server configured in Barman
* Added the 'recover' command which performs local recovery of a given
backup, allowing DBAs to specify a point in time. The 'recover' command
supports relocation of both the PGDATA directory and, where applicable,
the tablespaces
* Added the '--remote-ssh-command' option to the 'recover' command for
remote recovery of a backup. Remote recovery does not currently support
relocation of tablespaces
* Added the 'list-server' command that lists all the active servers
that have been configured in barman
* Added the 'show-server' command that shows the relevant information
for a given server, including all configuration options
* Added the 'status' command which shows information about the current
state of a server, including Postgres version, current transaction ID,
archive command, etc.
* Added the 'check' command which returns 0 if everything Barman needs
is functioning correctly
* Added the 'list-backup' command that lists all the available backups
for a given server, including size of the base backup and total size
of the related WAL segments
* Added the 'show-backup' command that shows the relevant information
for a given backup, including time of start, size, number of related
WAL segments and their size, etc.
* Added the 'delete' command which removes a backup from the catalog
* Added the 'list-files' command which lists all the files for a
single backup
* RPM Package for RHEL 5/6
barman-2.1/MANIFEST.in 0000664 0001750 0001750 00000000447 13020304015 014741 0 ustar gcalacoci gcalacoci recursive-include barman *.py
recursive-include rpm *
recursive-include doc *.md
recursive-include doc/barman.d *
include doc/Makefile doc/barman.1 doc/barman.5 doc/barman.conf
include scripts/barman.bash_completion
include AUTHORS NEWS ChangeLog LICENSE MANIFEST.in setup.py INSTALL README.rst
barman-2.1/scripts/ 0000775 0001750 0001750 00000000000 13033432341 014675 5 ustar gcalacoci gcalacoci barman-2.1/scripts/barman.bash_completion 0000664 0001750 0001750 00000000055 12726221047 021234 0 ustar gcalacoci gcalacoci eval "$(register-python-argcomplete barman)"
barman-2.1/rpm/ 0000775 0001750 0001750 00000000000 13033432341 014004 5 ustar gcalacoci gcalacoci barman-2.1/rpm/rhel7/ 0000775 0001750 0001750 00000000000 13033432341 015025 5 ustar gcalacoci gcalacoci barman-2.1/rpm/rhel7/python-argh.spec 0000664 0001750 0001750 00000003520 12726221047 020150 0 ustar gcalacoci gcalacoci %{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
Summary: A simple argparse wrapper
Name: python-argh
Version: 0.26.1
Release: 1%{?dist}
License: LGPLv3
Group: Development/Libraries
Url: http://bitbucket.org/neithere/argh/
Source0: http://pypi.python.org/packages/source/a/argh/argh-%{version}.tar.gz
BuildRequires: python-devel, python-setuptools
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot-%(%{__id_u} -n)
BuildArch: noarch
Requires: python-abi = %(%{__python} -c "import sys ; print sys.version[:3]")
Requires: python-argparse
%description
Argh, argparse!
===============
Did you ever say "argh" trying to remember the details of optparse or argparse
API? If yes, this package may be useful for you. It provides a very simple
wrapper for argparse with support for hierarchical commands that can be bound
to modules or classes. Argparse can do it; argh makes it easy.
%prep
%setup -n argh-%{version} -q
%build
%{__python} setup.py build
%install
%{__python} setup.py install -O1 --skip-build --root $RPM_BUILD_ROOT
%clean
rm -rf $RPM_BUILD_ROOT
%files
%defattr(-,root,root)
%doc README.rst
%{python_sitelib}/argh-%{version}-py2.7.egg-info
%{python_sitelib}/argh/
%changelog
* Tue Jan 20 2015 - Francesco Canovai 0.26.1-1
- Update to version 0.26.1
* Thu Jan 31 2013 - Marco Nenciarini 0.23.0-1
- Update to version 0.23.0
* Wed May 9 2012 - Marco Nenciarini 0.15.0-1
- Update to version 0.15.0
* Sat Dec 3 2011 - Marco Nenciarini 0.14.2-1
- Initial packaging.
barman-2.1/rpm/barman.spec 0000664 0001750 0001750 00000010550 13033425212 016120 0 ustar gcalacoci gcalacoci %if 0%{?rhel} == 7
%global pybasever 2.7
%else
%if 0%{?fedora}>=21
%global pybasever 2.7
%else
%global pybasever 2.6
%endif
%endif
%if 0%{?rhel} == 5
%global with_python26 1
%endif
%if 0%{?with_python26}
%global __python_ver python26
%global __python %{_bindir}/python%{pybasever}
%global __os_install_post %{__multiple_python_os_install_post}
%else
%global __python_ver python
%endif
%global main_version 2.1
# comment out the next line if not a pre-release (use '#%%global ...')
#%%global extra_version a1
# Usually 1 - unique sequence for all pre-release version
%global package_release 1
%{!?pybasever: %define pybasever %(%{__python} -c "import sys;print(sys.version[0:3])")}
%{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
Summary: Backup and Recovery Manager for PostgreSQL
Name: barman
Version: %{main_version}
Release: %{?extra_version:0.}%{package_release}%{?extra_version:.%{extra_version}}%{?dist}
License: GPLv3
Group: Applications/Databases
Url: http://www.pgbarman.org/
Source0: %{name}-%{version}%{?extra_version:%{extra_version}}.tar.gz
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot-%(%{__id_u} -n)
BuildArch: noarch
Vendor: 2ndQuadrant Italia Srl
Requires: python-abi = %{pybasever}, %{__python_ver}-psycopg2 >= 2.4.2, %{__python_ver}-argh >= 0.21.2, %{__python_ver}-argcomplete, %{__python_ver}-dateutil
Requires: /usr/sbin/useradd
Requires: rsync >= 3.0.4
%description
Barman (Backup and Recovery Manager) is an open-source
administration tool for disaster recovery of PostgreSQL
servers written in Python.
It allows your organisation to perform remote backups of
multiple servers in business critical environments to
reduce risk and help DBAs during the recovery phase.
Barman is distributed under GNU GPL 3 and maintained
by 2ndQuadrant.
%prep
%setup -n barman-%{version}%{?extra_version:%{extra_version}} -q
%build
%{__python} setup.py build
cat > barman.cron << EOF
# m h dom mon dow user command
* * * * * barman [ -x %{_bindir}/barman ] && %{_bindir}/barman -q cron
EOF
cat > barman.logrotate << EOF
/var/log/barman/barman.log {
missingok
notifempty
create 0600 barman barman
}
EOF
%install
%{__python} setup.py install -O1 --skip-build --root %{buildroot}
mkdir -p %{buildroot}%{_sysconfdir}/bash_completion.d
mkdir -p %{buildroot}%{_sysconfdir}/cron.d/
mkdir -p %{buildroot}%{_sysconfdir}/logrotate.d/
mkdir -p %{buildroot}%{_sysconfdir}/barman.d/
mkdir -p %{buildroot}/var/lib/barman
mkdir -p %{buildroot}/var/log/barman
install -pm 644 doc/barman.conf %{buildroot}%{_sysconfdir}/barman.conf
install -pm 644 doc/barman.d/* %{buildroot}%{_sysconfdir}/barman.d/
install -pm 644 scripts/barman.bash_completion %{buildroot}%{_sysconfdir}/bash_completion.d/barman
install -pm 644 barman.cron %{buildroot}%{_sysconfdir}/cron.d/barman
install -pm 644 barman.logrotate %{buildroot}%{_sysconfdir}/logrotate.d/barman
touch %{buildroot}/var/log/barman/barman.log
%clean
rm -rf %{buildroot}
%files
%defattr(-,root,root)
%doc INSTALL NEWS README.rst
%{python_sitelib}/%{name}-%{version}%{?extra_version:%{extra_version}}-py%{pybasever}.egg-info
%{python_sitelib}/%{name}/
%{_bindir}/%{name}
%doc %{_mandir}/man1/%{name}.1.gz
%doc %{_mandir}/man5/%{name}.5.gz
%config(noreplace) %{_sysconfdir}/bash_completion.d/
%config(noreplace) %{_sysconfdir}/%{name}.conf
%config(noreplace) %{_sysconfdir}/cron.d/%{name}
%config(noreplace) %{_sysconfdir}/logrotate.d/%{name}
%config(noreplace) %{_sysconfdir}/barman.d/
%attr(700,barman,barman) %dir /var/lib/%{name}
%attr(755,barman,barman) %dir /var/log/%{name}
%attr(600,barman,barman) %ghost /var/log/%{name}/%{name}.log
%pre
groupadd -f -r barman >/dev/null 2>&1 || :
useradd -M -n -g barman -r -d /var/lib/barman -s /bin/bash \
-c "Backup and Recovery Manager for PostgreSQL" barman >/dev/null 2>&1 || :
%changelog
* Thu Jan 5 2017 - Giulio Calacoci 2.1-1
- New release 2.1-1
* Tue Dec 27 2016 - Gabriele Bartolini 2.1-0.1.alpha.1
- New release 2.1-0.1.alpha.1
* Tue Sep 27 2016 - Gabriele Bartolini 2.0-1
- New release 2.0-1
- Trim changelog for releases 1.X
barman-2.1/rpm/rhel6/ 0000775 0001750 0001750 00000000000 13033432341 015024 5 ustar gcalacoci gcalacoci barman-2.1/rpm/rhel6/python-argh.spec 0000664 0001750 0001750 00000003332 12726221047 020150 0 ustar gcalacoci gcalacoci %{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
Summary: A simple argparse wrapper
Name: python-argh
Version: 0.23.0
Release: 1%{?dist}
License: LGPLv3
Group: Development/Libraries
Url: http://bitbucket.org/neithere/argh/
Source0: http://pypi.python.org/packages/source/a/argh/argh-%{version}.tar.gz
BuildRequires: python-devel, python-setuptools
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot-%(%{__id_u} -n)
BuildArch: noarch
Requires: python-abi = %(%{__python} -c "import sys ; print sys.version[:3]")
Requires: python-argparse
%description
Argh, argparse!
===============
Did you ever say "argh" trying to remember the details of optparse or argparse
API? If yes, this package may be useful for you. It provides a very simple
wrapper for argparse with support for hierarchical commands that can be bound
to modules or classes. Argparse can do it; argh makes it easy.
%prep
%setup -n argh-%{version} -q
%build
%{__python} setup.py build
%install
%{__python} setup.py install -O1 --skip-build --root $RPM_BUILD_ROOT
%clean
rm -rf $RPM_BUILD_ROOT
%files
%defattr(-,root,root)
%doc README
%{python_sitelib}/argh-%{version}-py2.6.egg-info
%{python_sitelib}/argh/
%changelog
* Thu Jan 31 2013 - Marco Neciarini 0.23.0-1
- Update to version 0.23.0
* Wed May 9 2012 - Marco Neciarini 0.15.0-1
- Update to version 0.15.0
* Sat Dec 4 2011 - Marco Neciarini 0.14.2-1
- Initial packaging.
barman-2.1/rpm/rhel6/python-argcomplete.spec 0000664 0001750 0001750 00000003537 12726221047 021540 0 ustar gcalacoci gcalacoci %{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
Summary: Bash tab completion for argparse
Name: python-argcomplete
Version: 0.3.5
Release: 1%{?dist}
License: ASL 2.0
Group: Development/Libraries
Url: https://github.com/kislyuk/argcomplete
Source0: http://pypi.python.org/packages/source/a/argcomplete/argcomplete-%{version}.tar.gz
BuildRequires: python-devel,python-setuptools
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot-%(%{__id_u} -n)
BuildArch: noarch
Requires: python-abi = %(%{__python} -c "import sys ; print sys.version[:3]")
Requires: python-argparse
%description
Argcomplete provides easy, extensible command line tab completion of
arguments for your Python script.
It makes two assumptions:
* You're using bash as your shell
* You're using argparse to manage your command line arguments/options
Argcomplete is particularly useful if your program has lots of
options or subparsers, and if your program can dynamically suggest
completions for your argument/option values (for example, if the user
is browsing resources over the network).
%prep
%setup -n argcomplete-%{version} -q
%build
%{__python} setup.py build
%install
%{__python} setup.py install -O1 --skip-build --root $RPM_BUILD_ROOT
%clean
rm -rf $RPM_BUILD_ROOT
%files
%defattr(-,root,root)
%doc README.rst
%{python_sitelib}/argcomplete-%{version}-py2.6.egg-info
%{python_sitelib}/argcomplete/
%{_bindir}/activate-global-python-argcomplete
%{_bindir}/python-argcomplete-check-easy-install-script
%{_bindir}/register-python-argcomplete
%changelog
* Thu Jan 31 2013 - Marco Neciarini 0.3.5-1
- Initial packaging.
barman-2.1/rpm/rhel5/ 0000775 0001750 0001750 00000000000 13033432341 015023 5 ustar gcalacoci gcalacoci barman-2.1/rpm/rhel5/python26-psycopg2.spec 0000644 0001750 0001750 00000014000 12402340601 021121 0 ustar gcalacoci gcalacoci # Use Python 2.6
%global pybasever 2.6
%global __python_ver 26
%global __python %{_bindir}/python%{pybasever}
%global __os_install_post %{__multiple_python_os_install_post}
%{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
%define ZPsycopgDAdir %{_localstatedir}/lib/zope/Products/ZPsycopgDA
%global pgmajorversion 90
%global pginstdir /usr/pgsql-9.0
%global sname psycopg2
Summary: A PostgreSQL database adapter for Python
Name: python26-%{sname}
Version: 2.4.5
Release: 1%{?dist}
License: LGPLv3 with exceptions
Group: Applications/Databases
Url: http://www.psycopg.org/psycopg/
Source0: http://initd.org/psycopg/tarballs/PSYCOPG-2-4/%{sname}-%{version}.tar.gz
Patch0: setup.cfg.patch
BuildRequires: python%{__python_ver}-devel postgresql%{pgmajorversion}-devel
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot-%(%{__id_u} -n)
Requires: python-abi = %(%{__python} -c "import sys ; print sys.version[:3]")
%description
psycopg is a PostgreSQL database adapter for the Python programming
language (just like pygresql and popy.) It was written from scratch
with the aim of being very small and fast, and stable as a rock. The
main advantages of psycopg are that it supports the full Python
DBAPI-2.0 and being thread safe at level 2.
%package doc
Summary: Documentation for psycopg python PostgreSQL database adapter
Group: Documentation
Requires: %{name} = %{version}-%{release}
%description doc
Documentation and example files for the psycopg python PostgreSQL
database adapter.
%package test
Summary: Tests for psycopg2
Group: Development/Libraries
Requires: %{name} = %{version}-%{release}
%description test
Tests for psycopg2.
%package zope
Summary: Zope Database Adapter ZPsycopgDA
Group: Applications/Databases
Requires: %{name} = %{version}-%{release} zope
%description zope
Zope Database Adapter for PostgreSQL, called ZPsycopgDA
%prep
%setup -q -n psycopg2-%{version}
%patch0 -p0
%build
%{__python} setup.py build
# Fix for wrong-file-end-of-line-encoding problem; upstream also must fix this.
for i in `find doc -iname "*.html"`; do sed -i 's/\r//' $i; done
for i in `find doc -iname "*.css"`; do sed -i 's/\r//' $i; done
%install
rm -Rf %{buildroot}
mkdir -p %{buildroot}%{python_sitearch}/psycopg2
%{__python} setup.py install --no-compile --root %{buildroot}
install -d %{buildroot}%{ZPsycopgDAdir}
cp -pr ZPsycopgDA/* %{buildroot}%{ZPsycopgDAdir}
%clean
rm -rf %{buildroot}
%files
%defattr(-,root,root)
%doc AUTHORS ChangeLog INSTALL LICENSE README
%dir %{python_sitearch}/psycopg2
%{python_sitearch}/psycopg2/*.py
%{python_sitearch}/psycopg2/*.pyc
%{python_sitearch}/psycopg2/*.so
%{python_sitearch}/psycopg2/*.pyo
%{python_sitearch}/psycopg2-*.egg-info
%files doc
%defattr(-,root,root)
%doc doc examples/
%files test
%defattr(-,root,root)
%{python_sitearch}/%{sname}/tests/*
%files zope
%defattr(-,root,root)
%dir %{ZPsycopgDAdir}
%{ZPsycopgDAdir}/*.py
%{ZPsycopgDAdir}/*.pyo
%{ZPsycopgDAdir}/*.pyc
%{ZPsycopgDAdir}/dtml/*
%{ZPsycopgDAdir}/icons/*
%changelog
* Wed May 9 2012 - Marco Neciarini 2.4.5-1
- Update to version 2.4.5
* Mon Aug 22 2011 Devrim GUNDUZ 2.4.2-1
- Update to 2.4.2
- Add a patch for pg_config path.
- Add new subpackage: test
* Tue Mar 16 2010 Devrim GUNDUZ 2.0.14-1
- Update to 2.0.14
* Mon Oct 19 2009 Devrim GUNDUZ 2.0.13-1
- Update to 2.0.13
* Mon Sep 7 2009 Devrim GUNDUZ 2.0.12-1
- Update to 2.0.12
* Tue May 26 2009 Devrim GUNDUZ 2.0.11-1
- Update to 2.0.11
* Fri Apr 24 2009 Devrim GUNDUZ 2.0.10-1
- Update to 2.0.10
* Thu Mar 2 2009 Devrim GUNDUZ 2.0.9-1
- Update to 2.0.9
* Wed Apr 30 2008 - Devrim GUNDUZ 2.0.7-1
- Update to 2.0.7
* Fri Jun 15 2007 - Devrim GUNDUZ 2.0.6-1
- Update to 2.0.6
* Sun May 06 2007 Thorsten Leemhuis
- rebuilt for RHEL5 final
* Wed Dec 6 2006 - Devrim GUNDUZ 2.0.5.1-4
- Rebuilt for PostgreSQL 8.2.0
* Mon Sep 11 2006 - Devrim GUNDUZ 2.0.5.1-3
- Rebuilt
* Wed Sep 6 2006 - Devrim GUNDUZ 2.0.5.1-2
- Remove ghost'ing, per Python Packaging Guidelines
* Mon Sep 4 2006 - Devrim GUNDUZ 2.0.5.1-1
- Update to 2.0.5.1
* Sun Aug 6 2006 - Devrim GUNDUZ 2.0.3-3
- Fixed zope package dependencies and macro definition, per bugzilla review (#199784)
- Fixed zope package directory ownership, per bugzilla review (#199784)
- Fixed cp usage for zope subpackage, per bugzilla review (#199784)
* Mon Jul 31 2006 - Devrim GUNDUZ 2.0.3-2
- Fixed 64 bit builds
- Fixed license
- Added Zope subpackage
- Fixed typo in doc description
- Added macro for zope subpackage dir
* Mon Jul 31 2006 - Devrim GUNDUZ 2.0.3-1
- Update to 2.0.3
- Fixed spec file, per bugzilla review (#199784)
* Sat Jul 22 2006 - Devrim GUNDUZ 2.0.2-3
- Removed python dependency, per bugzilla review. (#199784)
- Changed doc package group, per bugzilla review. (#199784)
- Replaced dos2unix with sed, per guidelines and bugzilla review (#199784)
- Fix changelog dates
* Sat Jul 21 2006 - Devrim GUNDUZ 2.0.2-2
- Added dos2unix to buildrequires
- removed python related part from package name
* Fri Jul 20 2006 - Devrim GUNDUZ 2.0.2-1
- Fix rpmlint errors, including dos2unix solution
- Re-engineered spec file
* Fri Jan 23 2006 - Devrim GUNDUZ
- First 2.0.X build
* Fri Jan 23 2006 - Devrim GUNDUZ
- Update to 1.2.21
* Tue Dec 06 2005 - Devrim GUNDUZ
- Initial release for 1.1.20
barman-2.1/rpm/rhel5/python26-dateutil.spec 0000644 0001750 0001750 00000010144 12402340601 021173 0 ustar gcalacoci gcalacoci # Use Python 2.6
%global pybasever 2.6
%global __python_ver 26
%global __python %{_bindir}/python%{pybasever}
%global __os_install_post %{__multiple_python_os_install_post}
%{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
Name: python%{__python_ver}-dateutil
Version: 1.4.1
Release: 6%{?dist}
Summary: Powerful extensions to the standard datetime module
Group: Development/Languages
License: Python
URL: http://labix.org/python-dateutil
Source0: http://labix.org/download/python-dateutil/python-dateutil-%{version}.tar.gz
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n)
# Redirect the exposed parts of the dateutil.zoneinfo API to remove references
# to the embedded copy of zoneinfo-2008e.tar.gz and instead use the system
# data from the "tzdata" package (rhbz#559309):
Patch0: python-dateutil-1.4.1-remove-embedded-timezone-data.patch
BuildArch: noarch
BuildRequires: python%{__python_ver}-devel,python%{__python_ver}-setuptools
Requires: tzdata
%description
The dateutil module provides powerful extensions to the standard datetime
module available in Python 2.3+.
%prep
%setup -n python-dateutil-%{version} -q
# Remove embedded copy of timezone data:
%patch0 -p1
rm dateutil/zoneinfo/zoneinfo-2008e.tar.gz
# Change encoding of NEWS file to UTF-8, preserving timestamp:
iconv -f ISO-8859-1 -t utf8 NEWS > NEWS.utf8 && \
touch -r NEWS NEWS.utf8 && \
mv NEWS.utf8 NEWS
%build
%{__python} setup.py build
%install
rm -rf $RPM_BUILD_ROOT
%{__python} setup.py install -O1 --skip-build --root $RPM_BUILD_ROOT
%clean
rm -rf $RPM_BUILD_ROOT
%check
%{__python} test.py
%files
%defattr(-,root,root,-)
%doc example.py LICENSE NEWS README
%{python_sitelib}/dateutil/
%{python_sitelib}/*.egg-info
%changelog
* Tue Jul 13 2010 David Malcolm - 1.4.1-6
- remove embedded copy of timezone data, and redirect the dateutil.zoneinfo
API accordingly
Resolves: rhbz#559309
- add a %%check, running the upstream selftest suite
* Tue Jul 13 2010 David Malcolm - 1.4.1-5
- add requirement on tzdata
Resolves: rhbz#559309
- fix encoding of the NEWS file
* Mon Nov 30 2009 Dennis Gregorovic - 1.4.1-4.1
- Rebuilt for RHEL 6
* Sun Jul 26 2009 Fedora Release Engineering - 1.4.1-4
- Rebuilt for https://fedoraproject.org/wiki/Fedora_12_Mass_Rebuild
* Thu Feb 26 2009 Fedora Release Engineering - 1.4.1-3
- Rebuilt for https://fedoraproject.org/wiki/Fedora_11_Mass_Rebuild
* Fri Feb 20 2009 Jef Spaleta - 1.4.1-2
- small specfile fix
* Fri Feb 20 2009 Jef Spaleta - 1.4.1-2
- New upstream version
* Sat Nov 29 2008 Ignacio Vazquez-Abrams - 1.4-3
- Rebuild for Python 2.6
* Fri Aug 29 2008 Tom "spot" Callaway - 1.4-2
- fix license tag
* Tue Jul 01 2008 Jef Spaleta 1.4-1
- Latest upstream release
* Fri Jan 04 2008 Jef Spaleta 1.2-2
- Fix for egg-info file creation
* Thu Jun 28 2007 Orion Poplawski 1.2-1
- Update to 1.2
* Mon Dec 11 2006 Jef Spaleta 1.1-5
- Fix python-devel BR, as per discussion in maintainers-list
* Mon Dec 11 2006 Jef Spaleta 1.1-4
- Release bump for rebuild against python 2.5 in devel tree
* Wed Jul 26 2006 Orion Poplawski 1.1-3
- Add patch to fix building on x86_64
* Wed Feb 15 2006 Orion Poplawski 1.1-2
- Rebuild for gcc/glibc changes
* Thu Dec 22 2005 Orion Poplawski 1.1-1
- Update to 1.1
* Thu Jul 28 2005 Orion Poplawski 1.0-1
- Update to 1.0
* Tue Jul 05 2005 Orion Poplawski 0.9-1
- Initial Fedora Extras package
barman-2.1/rpm/rhel5/python26-argh.spec 0000664 0001750 0001750 00000003762 12726221047 020326 0 ustar gcalacoci gcalacoci # Use Python 2.6
%global pybasever 2.6
%global __python_ver 26
%global __python %{_bindir}/python%{pybasever}
%global __os_install_post %{__multiple_python_os_install_post}
%{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
Summary: A simple argparse wrapper
Name: python%{__python_ver}-argh
Version: 0.23.0
Release: 1%{?dist}
License: LGPLv3
Group: Development/Libraries
Url: http://bitbucket.org/neithere/argh/
Source0: http://pypi.python.org/packages/source/a/argh/argh-%{version}.tar.gz
BuildRequires: python%{__python_ver}-devel,python%{__python_ver}-setuptools
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot-%(%{__id_u} -n)
BuildArch: noarch
Requires: python-abi = %(%{__python} -c "import sys ; print sys.version[:3]")
%if "%{__python_ver}" == "26"
Requires: python%{__python_ver}-argparse
%endif
%description
Agrh, argparse!
===============
Did you ever say "argh" trying to remember the details of optparse or argparse
API? If yes, this package may be useful for you. It provides a very simple
wrapper for argparse with support for hierarchical commands that can be bound
to modules or classes. Argparse can do it; argh makes it easy.
%prep
%setup -n argh-%{version} -q
%build
%{__python} setup.py build
%install
%{__python} setup.py install -O1 --skip-build --root $RPM_BUILD_ROOT
%clean
rm -rf $RPM_BUILD_ROOT
%files
%defattr(-,root,root)
%doc README
%{python_sitelib}/argh-%{version}-py%{pybasever}.egg-info
%{python_sitelib}/argh/
%changelog
* Thu Jan 31 2013 - Marco Neciarini 0.23.0-1
- Update to version 0.23.0
* Wed May 9 2012 - Marco Neciarini 0.15.0-1
- Update to version 0.15.0
* Sat Dec 4 2011 - Marco Neciarini 0.14.2-1
- Initial packaging.
barman-2.1/rpm/rhel5/python-dateutil-1.4.1-remove-embedded-timezone-data.patch 0000644 0001750 0001750 00000006367 12402340601 027424 0 ustar gcalacoci gcalacoci diff -up python-dateutil-1.4.1/dateutil/tz.py.remove-embedded-timezone-data python-dateutil-1.4.1/dateutil/tz.py
--- python-dateutil-1.4.1/dateutil/tz.py.remove-embedded-timezone-data 2008-02-27 20:45:41.000000000 -0500
+++ python-dateutil-1.4.1/dateutil/tz.py 2010-07-13 14:40:30.228122861 -0400
@@ -930,9 +930,6 @@ def gettz(name=None):
except OSError:
pass
if not tz:
- from dateutil.zoneinfo import gettz
- tz = gettz(name)
- if not tz:
for c in name:
# name must have at least one offset to be a tzstr
if c in "0123456789":
diff -up python-dateutil-1.4.1/dateutil/zoneinfo/__init__.py.remove-embedded-timezone-data python-dateutil-1.4.1/dateutil/zoneinfo/__init__.py
--- python-dateutil-1.4.1/dateutil/zoneinfo/__init__.py.remove-embedded-timezone-data 2005-12-22 13:13:50.000000000 -0500
+++ python-dateutil-1.4.1/dateutil/zoneinfo/__init__.py 2010-07-13 14:40:30.228122861 -0400
@@ -3,6 +3,10 @@ Copyright (c) 2003-2005 Gustavo Niemeye
This module offers extensions to the standard python 2.3+
datetime module.
+
+This version of the code has been modified to remove the embedded copy
+of zoneinfo-2008e.tar.gz and instead use the system data from the tzdata
+package
"""
from dateutil.tz import tzfile
from tarfile import TarFile
@@ -13,49 +17,12 @@ __license__ = "PSF License"
__all__ = ["setcachesize", "gettz", "rebuild"]
-CACHE = []
-CACHESIZE = 10
-
-class tzfile(tzfile):
- def __reduce__(self):
- return (gettz, (self._filename,))
-
-def getzoneinfofile():
- filenames = os.listdir(os.path.join(os.path.dirname(__file__)))
- filenames.sort()
- filenames.reverse()
- for entry in filenames:
- if entry.startswith("zoneinfo") and ".tar." in entry:
- return os.path.join(os.path.dirname(__file__), entry)
- return None
-
-ZONEINFOFILE = getzoneinfofile()
-
-del getzoneinfofile
-
def setcachesize(size):
- global CACHESIZE, CACHE
- CACHESIZE = size
- del CACHE[size:]
+ pass
def gettz(name):
- tzinfo = None
- if ZONEINFOFILE:
- for cachedname, tzinfo in CACHE:
- if cachedname == name:
- break
- else:
- tf = TarFile.open(ZONEINFOFILE)
- try:
- zonefile = tf.extractfile(name)
- except KeyError:
- tzinfo = None
- else:
- tzinfo = tzfile(zonefile)
- tf.close()
- CACHE.insert(0, (name, tzinfo))
- del CACHE[CACHESIZE:]
- return tzinfo
+ from dateutil.tz import gettz
+ return gettz(name)
def rebuild(filename, tag=None, format="gz"):
import tempfile, shutil
diff -up python-dateutil-1.4.1/MANIFEST.in.remove-embedded-timezone-data python-dateutil-1.4.1/MANIFEST.in
--- python-dateutil-1.4.1/MANIFEST.in.remove-embedded-timezone-data 2010-07-13 14:42:07.974118722 -0400
+++ python-dateutil-1.4.1/MANIFEST.in 2010-07-13 14:42:14.409994960 -0400
@@ -1,4 +1,4 @@
-recursive-include dateutil *.py *.tar.*
+recursive-include dateutil *.py
recursive-include sandbox *.py
include setup.py setup.cfg MANIFEST.in README LICENSE NEWS Makefile
include test.py example.py
barman-2.1/rpm/rhel5/setup.cfg.patch 0000644 0001750 0001750 00000001001 12402340601 017724 0 ustar gcalacoci gcalacoci --- setup.cfg.old 2011-08-22 12:16:18.703486005 +0300
+++ setup.cfg 2011-08-22 12:16:31.596486005 +0300
@@ -26,7 +26,7 @@
# libraries needed to build psycopg2. If pg_config is not in the path or
# is installed under a different name uncomment the following option and
# set it to the pg_config full path.
-#pg_config=
+pg_config=/usr/pgsql-9.0/bin/pg_config
# If "pg_config" is not available, "include_dirs" can be used to locate
# postgresql headers and libraries. Some extra checks on sys.platform will
barman-2.1/rpm/rhel5/python26-argcomplete.spec 0000664 0001750 0001750 00000004170 12726221047 021701 0 ustar gcalacoci gcalacoci # Use Python 2.6
%global pybasever 2.6
%global __python_ver 26
%global __python %{_bindir}/python%{pybasever}
%global __os_install_post %{__multiple_python_os_install_post}
%{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")}
%{!?python_sitearch: %define python_sitearch %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib(1)")}
Summary: Bash tab completion for argparse
Name: python%{__python_ver}-argcomplete
Version: 0.3.5
Release: 1%{?dist}
License: ASL 2.0
Group: Development/Libraries
Url: https://github.com/kislyuk/argcomplete
Source0: http://pypi.python.org/packages/source/a/argcomplete/argcomplete-%{version}.tar.gz
BuildRequires: python%{__python_ver}-devel,python%{__python_ver}-setuptools
BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot-%(%{__id_u} -n)
BuildArch: noarch
Requires: python-abi = %(%{__python} -c "import sys ; print sys.version[:3]")
%if "%{__python_ver}" == "26"
Requires: python%{__python_ver}-argparse
%endif
%description
Argcomplete provides easy, extensible command line tab completion of
arguments for your Python script.
It makes two assumptions:
* You're using bash as your shell
* You're using argparse to manage your command line arguments/options
Argcomplete is particularly useful if your program has lots of
options or subparsers, and if your program can dynamically suggest
completions for your argument/option values (for example, if the user
is browsing resources over the network).
%prep
%setup -n argcomplete-%{version} -q
%build
%{__python} setup.py build
%install
%{__python} setup.py install -O1 --skip-build --root $RPM_BUILD_ROOT
%clean
rm -rf $RPM_BUILD_ROOT
%files
%defattr(-,root,root)
%doc README.rst
%{python_sitelib}/argcomplete-%{version}-py%{pybasever}.egg-info
%{python_sitelib}/argcomplete/
%{_bindir}/activate-global-python-argcomplete
%{_bindir}/python-argcomplete-check-easy-install-script
%{_bindir}/register-python-argcomplete
%changelog
* Thu Jan 31 2013 - Marco Neciarini 0.3.5-1
- Initial packaging.
barman-2.1/bin/ 0000775 0001750 0001750 00000000000 13033432341 013756 5 ustar gcalacoci gcalacoci barman-2.1/bin/barman 0000775 0001750 0001750 00000001526 12726221047 015157 0 ustar gcalacoci gcalacoci #!/usr/bin/env python
#
# Copyright (C) 2011-2016 2ndQuadrant Italia Srl
#
# This file is part of Barman.
#
# Barman is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Barman is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Barman. If not, see .
#
# PYTHON_ARGCOMPLETE_OK
from barman.cli import main
if __name__ == '__main__':
main()
else:
raise NotImplementedError
barman-2.1/doc/ 0000775 0001750 0001750 00000000000 13033432341 013753 5 ustar gcalacoci gcalacoci barman-2.1/doc/barman.conf 0000664 0001750 0001750 00000005225 13015546105 016072 0 ustar gcalacoci gcalacoci ; Barman, Backup and Recovery Manager for PostgreSQL
; http://www.pgbarman.org/ - http://www.2ndQuadrant.com/
;
; Main configuration file
[barman]
; System user
barman_user = barman
; Directory of configuration files. Place your sections in separate files with .conf extension
; For example place the 'main' server section in /etc/barman.d/main.conf
configuration_files_directory = /etc/barman.d
; Main directory
barman_home = /var/lib/barman
; Locks directory - default: %(barman_home)s
;barman_lock_directory = /var/run/barman
; Log location
log_file = /var/log/barman/barman.log
; Log level (see https://docs.python.org/3/library/logging.html#levels)
log_level = INFO
; Default compression level: possible values are None (default), bzip2, gzip, pigz, pygzip or pybzip2
;compression = gzip
; Pre/post backup hook scripts
;pre_backup_script = env | grep ^BARMAN
;pre_backup_retry_script = env | grep ^BARMAN
;post_backup_retry_script = env | grep ^BARMAN
;post_backup_script = env | grep ^BARMAN
; Pre/post archive hook scripts
;pre_archive_script = env | grep ^BARMAN
;pre_archive_retry_script = env | grep ^BARMAN
;post_archive_retry_script = env | grep ^BARMAN
;post_archive_script = env | grep ^BARMAN
; Global retention policy (REDUNDANCY or RECOVERY WINDOW) - default empty
;retention_policy =
; Global bandwidth limit in KBPS - default 0 (meaning no limit)
;bandwidth_limit = 4000
; Immediate checkpoint for backup command - default false
;immediate_checkpoint = false
; Enable network compression for data transfers - default false
;network_compression = false
; Number of retries of data copy during base backup after an error - default 0
;basebackup_retry_times = 0
; Number of seconds of wait after a failed copy, before retrying - default 30
;basebackup_retry_sleep = 30
; Maximum execution time, in seconds, per server
; for a barman check command - default 30
;check_timeout = 30
; Time frame that must contain the latest backup date.
; If the latest backup is older than the time frame, barman check
; command will report an error to the user.
; If empty, the latest backup is always considered valid.
; Syntax for this option is: "i (DAYS | WEEKS | MONTHS)" where i is an
; integer > 0 which identifies the number of days | weeks | months of
; validity of the latest backup for this check. Also known as 'smelly backup'.
;last_backup_maximum_age =
; Minimum number of required backups (redundancy)
;minimum_redundancy = 1
; Examples of retention policies
; Retention policy (disabled)
;retention_policy =
; Retention policy (based on redundancy)
;retention_policy = REDUNDANCY 2
; Retention policy (based on recovery window)
;retention_policy = RECOVERY WINDOW OF 4 WEEKS
barman-2.1/doc/barman.5 0000664 0001750 0001750 00000043665 13033425212 015316 0 ustar gcalacoci gcalacoci .TH "BARMAN" "5" "January 5, 2017" "Barman User manuals" "Version 2.1"
.SH NAME
.PP
barman \- Backup and Recovery Manager for PostgreSQL
.SH DESCRIPTION
.PP
Barman is an administration tool for disaster recovery of PostgreSQL
servers written in Python and maintained by 2ndQuadrant.
Barman can perform remote backups of multiple servers in business
critical environments and helps DBAs during the recovery phase.
.SH CONFIGURATION FILE LOCATIONS
.PP
The system\-level Barman configuration file is located at
.IP
.nf
\f[C]
/etc/barman.conf
\f[]
.fi
.PP
or
.IP
.nf
\f[C]
/etc/barman/barman.conf
\f[]
.fi
.PP
and is overridden on a per\-user level by
.IP
.nf
\f[C]
$HOME/.barman.conf
\f[]
.fi
.SH CONFIGURATION FILE SYNTAX
.PP
The Barman configuration file is a plain \f[C]INI\f[] file.
There is a general section called \f[C][barman]\f[] and a section
\f[C][servername]\f[] for each server you want to backup.
Rows starting with \f[C];\f[] are comments.
.SH CONFIGURATION FILE DIRECTORY
.PP
Barman supports the inclusion of multiple configuration files, through
the \f[C]configuration_files_directory\f[] option.
Included files must contain only server specifications, not global
configurations.
If the value of \f[C]configuration_files_directory\f[] is a directory,
Barman reads all files with \f[C]\&.conf\f[] extension that exist in
that folder.
For example, if you set it to \f[C]/etc/barman.d\f[], you can specify
your PostgreSQL servers placing each section in a separate
\f[C]\&.conf\f[] file inside the \f[C]/etc/barman.d\f[] folder.
.SH OPTIONS
.TP
.B active
When set to \f[C]true\f[] (default), the server is in full operational
state.
When set to \f[C]false\f[], the server can be used for diagnostics, but
any operational command such as backup execution or WAL archiving is
temporarily disabled.
Setting \f[C]active=false\f[] is a good practice when adding a new node
to Barman.
Server.
.RS
.RE
.TP
.B archiver
This option allows you to activate log file shipping through
PostgreSQL\[aq]s \f[C]archive_command\f[] for a server.
If set to \f[C]true\f[] (default), Barman expects that continuous
archiving for a server is in place and will activate checks as well as
management (including compression) of WAL files that Postgres deposits
in the \f[I]incoming\f[] directory.
Setting it to \f[C]false\f[], will disable standard continuous archiving
for a server.
Global/Server.
(NOTE: this option is currently required to be enabled until Barman
natively supports physical replication slots.
Setting it to \f[C]false\f[] will result in a disabled server).
.RS
.RE
.TP
.B archiver_batch_size
This option allows you to activate batch processing of WAL files for the
\f[C]archiver\f[] process, by setting it to a value > 0.
Otherwise, the traditional unlimited processing of the WAL queue is
enabled.
When batch processing is activated, the \f[C]archive\-wal\f[] process
would limit itself to maximum \f[C]archiver_batch_size\f[] WAL segments
per single run.
Integer.
Global/Server.
.RS
.RE
.TP
.B backup_directory
Directory where backup data for a server will be placed.
Server.
.RS
.RE
.TP
.B backup_method
Configure the method barman used for backup execution.
If set to \f[C]rsync\f[] (default), barman will execute backup using the
\f[C]rsync\f[] command.
If set to \f[C]postgres\f[] barman will use the \f[C]pg_basebackup\f[]
command to execute the backup.
Global/Server.
.RS
.RE
.TP
.B backup_options
This option allows you to control the way Barman interacts with
PostgreSQL for backups.
If set to \f[C]exclusive_backup\f[] (default), \f[C]barman\ backup\f[]
executes backup operations using the standard exclusive backup approach
(technically through pg_start_backup/pg_stop_backup).
If set to \f[C]concurrent_backup\f[], if using PostgreSQL 9.2, 9.3, 9.4,
and 9.5, Barman requires the \f[C]pgespresso\f[] module to be installed
on the PostgreSQL server and can be used to perform a backup from a
standby server.
Starting from PostgreSQL 9.6, Barman uses the new PostgreSQL API to
perform backups from a standby server.
Global/Server.
.RS
.RE
.TP
.B bandwidth_limit
This option allows you to specify a maximum transfer rate in kilobytes
per second.
A value of zero specifies no limit (default).
Global/Server.
.RS
.RE
.TP
.B barman_home
Main data directory for Barman.
Global.
.RS
.RE
.TP
.B barman_lock_directory
Directory for locks.
Default: \f[C]%(barman_home)s\f[].
Global.
.RS
.RE
.TP
.B basebackups_directory
Directory where base backups will be placed.
Server.
.RS
.RE
.TP
.B basebackup_retry_sleep
Number of seconds of wait after a failed copy, before retrying Used
during both backup and recovery operations.
Positive integer, default 30.
Global/Server.
.RS
.RE
.TP
.B basebackup_retry_times
Number of retries of base backup copy, after an error.
Used during both backup and recovery operations.
Positive integer, default 0.
Global/Server.
.RS
.RE
.TP
.B check_timeout
Maximum execution time, in seconds per server, for a barman check
command.
Set to 0 to disable the timeout.
Positive integer, default 30.
Global/Server.
.RS
.RE
.TP
.B compression
Standard compression algorithm applied to WAL files.
Possible values are: \f[C]gzip\f[] (requires \f[C]gzip\f[] to be
installed on the system), \f[C]bzip2\f[] (requires \f[C]bzip2\f[]),
\f[C]pigz\f[] (requires \f[C]pigz\f[]), \f[C]pygzip\f[] (Python\[aq]s
internal gzip compressor) and \f[C]pybzip2\f[] (Python\[aq]s internal
bzip2 compressor).
Global/Server.
.RS
.RE
.TP
.B conninfo
Connection string used by Barman to connect to the Postgres server.
This is a libpq connection string, consult the PostgreSQL
manual (https://www.postgresql.org/docs/current/static/libpq-connect.html#LIBPQ-CONNSTRING)
for more information.
Commonly used keys are: host, hostaddr, port, dbname, user, password.
Server.
.RS
.RE
.TP
.B custom_compression_filter
Customised compression algorithm applied to WAL files.
Global/Server.
.RS
.RE
.TP
.B custom_decompression_filter
Customised decompression algorithm applied to compressed WAL files; this
must match the compression algorithm.
Global/Server.
.RS
.RE
.TP
.B description
A human readable description of a server.
Server.
.RS
.RE
.TP
.B errors_directory
Directory that contains WAL files that contain an error; usually this is
related to a conflict with an existing WAL file (e.g.
a WAL file that has been archived after a streamed one).
.RS
.RE
.TP
.B immediate_checkpoint
This option allows you to control the way PostgreSQL handles checkpoint
at the start of the backup.
If set to \f[C]false\f[] (default), the I/O workload for the checkpoint
will be limited, according to the \f[C]checkpoint_completion_target\f[]
setting on the PostgreSQL server.
If set to \f[C]true\f[], an immediate checkpoint will be requested,
meaning that PostgreSQL will complete the checkpoint as soon as
possible.
Global/Server.
.RS
.RE
.TP
.B incoming_wals_directory
Directory where incoming WAL files are archived into.
Requires \f[C]archiver\f[] to be enabled.
Server.
.RS
.RE
.TP
.B last_backup_maximum_age
This option identifies a time frame that must contain the latest backup.
If the latest backup is older than the time frame, barman check command
will report an error to the user.
If empty (default), latest backup is always considered valid.
Syntax for this option is: "i (DAYS | WEEKS | MONTHS)" where i is a
integer greater than zero, representing the number of days | weeks |
months of the time frame.
Global/Server.
.RS
.RE
.TP
.B log_file
Location of Barman\[aq]s log file.
Global.
.RS
.RE
.TP
.B log_level
Level of logging (DEBUG, INFO, WARNING, ERROR, CRITICAL).
Global.
.RS
.RE
.TP
.B minimum_redundancy
Minimum number of backups to be retained.
Default 0.
Global/Server.
.RS
.RE
.TP
.B network_compression
This option allows you to enable data compression for network transfers.
If set to \f[C]false\f[] (default), no compression is used.
If set to \f[C]true\f[], compression is enabled, reducing network usage.
Global/Server.
.RS
.RE
.TP
.B path_prefix
One or more absolute paths, separated by colon, where Barman looks for
executable files.
The paths specified in \f[C]path_prefix\f[] are tried before the ones
specified in \f[C]PATH\f[] environment variable.
Global/server.
.RS
.RE
.TP
.B post_archive_retry_script
Hook script launched after a WAL file is archived by maintenance.
Being this a \f[I]retry\f[] hook script, Barman will retry the execution
of the script until this either returns a SUCCESS (0), an ABORT_CONTINUE
(62) or an ABORT_STOP (63) code.
In a post archive scenario, ABORT_STOP has currently the same effects as
ABORT_CONTINUE.
Global/Server.
.RS
.RE
.TP
.B post_archive_script
Hook script launched after a WAL file is archived by maintenance, after
\[aq]post_archive_retry_script\[aq].
Global/Server.
.RS
.RE
.TP
.B post_backup_retry_script
Hook script launched after a base backup.
Being this a \f[I]retry\f[] hook script, Barman will retry the execution
of the script until this either returns a SUCCESS (0), an ABORT_CONTINUE
(62) or an ABORT_STOP (63) code.
In a post backup scenario, ABORT_STOP has currently the same effects as
ABORT_CONTINUE.
Global/Server.
.RS
.RE
.TP
.B post_backup_script
Hook script launched after a base backup, after
\[aq]post_backup_retry_script\[aq].
Global/Server.
.RS
.RE
.TP
.B pre_archive_retry_script
Hook script launched before a WAL file is archived by maintenance, after
\[aq]pre_archive_script\[aq].
Being this a \f[I]retry\f[] hook script, Barman will retry the execution
of the script until this either returns a SUCCESS (0), an ABORT_CONTINUE
(62) or an ABORT_STOP (63) code.
Returning ABORT_STOP will propagate the failure at a higher level and
interrupt the WAL archiving operation.
Global/Server.
.RS
.RE
.TP
.B pre_archive_script
Hook script launched before a WAL file is archived by maintenance.
Global/Server.
.RS
.RE
.TP
.B pre_backup_retry_script
Hook script launched before a base backup, after
\[aq]pre_backup_script\[aq].
Being this a \f[I]retry\f[] hook script, Barman will retry the execution
of the script until this either returns a SUCCESS (0), an ABORT_CONTINUE
(62) or an ABORT_STOP (63) code.
Returning ABORT_STOP will propagate the failure at a higher level and
interrupt the backup operation.
Global/Server.
.RS
.RE
.TP
.B pre_backup_script
Hook script launched before a base backup.
Global/Server.
.RS
.RE
.TP
.B recovery_options
Options for recovery operations.
Currently only supports \f[C]get\-wal\f[].
\f[C]get\-wal\f[] activates generation of a basic
\f[C]restore_command\f[] in the resulting \f[C]recovery.conf\f[] file
that uses the \f[C]barman\ get\-wal\f[] command to fetch WAL files
directly from Barman\[aq]s archive of WALs.
Comma separated list of values, default empty.
Global/Server.
.RS
.RE
.TP
.B retention_policy
Policy for retention of periodic backups and archive logs.
If left empty, retention policies are not enforced.
For redundancy based retention policy use "REDUNDANCY i" (where i is an
integer > 0 and defines the number of backups to retain).
For recovery window retention policy use "RECOVERY WINDOW OF i DAYS" or
"RECOVERY WINDOW OF i WEEKS" or "RECOVERY WINDOW OF i MONTHS" where i is
a positive integer representing, specifically, the number of days, weeks
or months to retain your backups.
For more detailed information, refer to the official documentation.
Default value is empty.
Global/Server.
.RS
.RE
.TP
.B retention_policy_mode
Currently only "auto" is implemented.
Global/Server.
.RS
.RE
.TP
.B reuse_backup
This option controls incremental backup support.
Global/Server.
Possible values are: * \f[C]off\f[]: disabled (default); *
\f[C]copy\f[]: reuse the last available backup for a server and create a
copy of the unchanged files (reduce backup time); * \f[C]link\f[]: reuse
the last available backup for a server and create a hard link of the
unchanged files (reduce backup time and space).
Requires operating system and file system support for hard links.
.RS
.RE
.TP
.B slot_name
Physical replication slot to be used by the \f[C]receive\-wal\f[]
command when \f[C]streaming_archiver\f[] is set to \f[C]on\f[].
Requires PostgreSQL >= 9.4.
Global/Server.
Default: None (disabled).
.RS
.RE
.TP
.B streaming_archiver
This option allows you to use the PostgreSQL\[aq]s streaming protocol to
receive transaction logs from a server.
If set to \f[C]on\f[], Barman expects to find \f[C]pg_receivexlog\f[] in
the PATH (see \f[C]path_prefix\f[] option) and that streaming connection
for the server is working.
This activates connection checks as well as management (including
compression) of WAL files.
If set to \f[C]off\f[] (default) barman will rely only on continuous
archiving for a server WAL archive operations, eventually terminating
any running \f[C]pg_receivexlog\f[] for the server.
Global/Server.
.RS
.RE
.TP
.B streaming_archiver_batch_size
This option allows you to activate batch processing of WAL files for the
\f[C]streaming_archiver\f[] process, by setting it to a value > 0.
Otherwise, the traditional unlimited processing of the WAL queue is
enabled.
When batch processing is activated, the \f[C]archive\-wal\f[] process
would limit itself to maximum \f[C]streaming_archiver_batch_size\f[] WAL
segments per single run.
Integer.
Global/Server.
.RS
.RE
.TP
.B streaming_archiver_name
Identifier to be used as \f[C]application_name\f[] by the
\f[C]receive\-wal\f[] command.
Only available with \f[C]pg_receivexlog\f[] >= 9.3.
By default it is set to \f[C]barman_receive_wal\f[].
Global/Server.
.RS
.RE
.TP
.B streaming_backup_name
Identifier to be used as \f[C]application_name\f[] by the
\f[C]pg_basebackup\f[] command.
Only available with \f[C]pg_basebackup\f[] >= 9.3.
By default it is set to \f[C]barman_streaming_backup\f[].
Global/Server.
.RS
.RE
.TP
.B streaming_conninfo
Connection string used by Barman to connect to the Postgres server via
streaming replication protocol.
By default it is set to \f[C]conninfo\f[].
Server.
.RS
.RE
.TP
.B streaming_wals_directory
Directory where WAL files are streamed from the PostgreSQL server to
Barman.
Requires \f[C]streaming_archiver\f[] to be enabled.
Server.
.RS
.RE
.TP
.B ssh_command
Command used by Barman to login to the Postgres server via ssh.
Server.
.RS
.RE
.TP
.B tablespace_bandwidth_limit
This option allows you to specify a maximum transfer rate in kilobytes
per second, by specifying a comma separated list of tablespaces (pairs
TBNAME:BWLIMIT).
A value of zero specifies no limit (default).
Global/Server.
.RS
.RE
.TP
.B wal_retention_policy
Policy for retention of archive logs (WAL files).
Currently only "MAIN" is available.
Global/Server.
.RS
.RE
.TP
.B wals_directory
Directory which contains WAL files.
Server.
.RS
.RE
.SH HOOK SCRIPTS
.PP
The script definition is passed to a shell and can return any exit code.
.PP
The shell environment will contain the following variables:
.TP
.B \f[C]BARMAN_CONFIGURATION\f[]
configuration file used by barman
.RS
.RE
.TP
.B \f[C]BARMAN_ERROR\f[]
error message, if any (only for the \[aq]post\[aq] phase)
.RS
.RE
.TP
.B \f[C]BARMAN_PHASE\f[]
\[aq]pre\[aq] or \[aq]post\[aq]
.RS
.RE
.TP
.B \f[C]BARMAN_RETRY\f[]
\f[C]1\f[] if it is a \f[I]retry script\f[] (from 1.5.0), \f[C]0\f[] if
not
.RS
.RE
.TP
.B \f[C]BARMAN_SERVER\f[]
name of the server
.RS
.RE
.PP
Backup scripts specific variables:
.TP
.B \f[C]BARMAN_BACKUP_DIR\f[]
backup destination directory
.RS
.RE
.TP
.B \f[C]BARMAN_BACKUP_ID\f[]
ID of the backup
.RS
.RE
.TP
.B \f[C]BARMAN_PREVIOUS_ID\f[]
ID of the previous backup (if present)
.RS
.RE
.TP
.B \f[C]BARMAN_STATUS\f[]
status of the backup
.RS
.RE
.TP
.B \f[C]BARMAN_VERSION\f[]
version of Barman
.RS
.RE
.PP
Archive scripts specific variables:
.TP
.B \f[C]BARMAN_SEGMENT\f[]
name of the WAL file
.RS
.RE
.TP
.B \f[C]BARMAN_FILE\f[]
full path of the WAL file
.RS
.RE
.TP
.B \f[C]BARMAN_SIZE\f[]
size of the WAL file
.RS
.RE
.TP
.B \f[C]BARMAN_TIMESTAMP\f[]
WAL file timestamp
.RS
.RE
.TP
.B \f[C]BARMAN_COMPRESSION\f[]
type of compression used for the WAL file
.RS
.RE
.PP
Only in case of retry hook scripts, the exit code of the script is
checked by Barman.
Output of hook scripts is simply written in the log file.
.SH EXAMPLE
.PP
Here is an example of configuration file:
.IP
.nf
\f[C]
[barman]
;\ Main\ directory
barman_home\ =\ /var/lib/barman
;\ System\ user
barman_user\ =\ barman
;\ Log\ location
log_file\ =\ /var/log/barman/barman.log
;\ Default\ compression\ level
;compression\ =\ gzip
;\ Incremental\ backup
reuse_backup\ =\ link
;\ \[aq]main\[aq]\ PostgreSQL\ Server\ configuration
[main]
;\ Human\ readable\ description
description\ =\ \ "Main\ PostgreSQL\ Database"
;\ SSH\ options
ssh_command\ =\ ssh\ postgres\@pg
;\ PostgreSQL\ connection\ string
conninfo\ =\ host=pg\ user=postgres
;\ PostgreSQL\ streaming\ connection\ string
streaming_conninfo\ =\ host=pg\ user=postgres
;\ Minimum\ number\ of\ required\ backups\ (redundancy)
minimum_redundancy\ =\ 1
;\ Retention\ policy\ (based\ on\ redundancy)
retention_policy\ =\ REDUNDANCY\ 2
\f[]
.fi
.SH SEE ALSO
.PP
\f[C]barman\f[] (1).
.SH AUTHORS
.PP
In alphabetical order:
.IP \[bu] 2
Gabriele Bartolini (project leader)
.IP \[bu] 2
Jonathan Battiato (QA/testing)
.IP \[bu] 2
Stefano Bianucci (developer, intern from University of Florence)
.IP \[bu] 2
Giuseppe Broccolo (QA/testing)
.IP \[bu] 2
Giulio Calacoci (developer)
.IP \[bu] 2
Francesco Canovai (QA/testing)
.IP \[bu] 2
Leonardo Cecchi (developer)
.IP \[bu] 2
Gianni Ciolli (QA/testing)
.IP \[bu] 2
Britt Cole (documentation)
.IP \[bu] 2
Marco Nenciarini (lead developer)
.IP \[bu] 2
Rubens Souza (QA/testing)
.PP
Past contributors:
.IP \[bu] 2
Carlo Ascani
.SH RESOURCES
.IP \[bu] 2
Homepage:
.IP \[bu] 2
Documentation:
.IP \[bu] 2
Professional support:
.SH COPYING
.PP
Barman is the exclusive property of 2ndQuadrant Italia and its code is
distributed under GNU General Public License v3.
.PP
Copyright (C) 2011\-2016 2ndQuadrant Italia Srl \-
http://www.2ndQuadrant.it/.
.SH AUTHORS
2ndQuadrant Italy .
barman-2.1/doc/barman.5.md 0000664 0001750 0001750 00000041314 13033425212 015702 0 ustar gcalacoci gcalacoci % BARMAN(5) Barman User manuals | Version 2.1
% 2ndQuadrant Italy
% January 5, 2017
# NAME
barman - Backup and Recovery Manager for PostgreSQL
# DESCRIPTION
Barman is an administration tool for disaster recovery of PostgreSQL
servers written in Python and maintained by 2ndQuadrant.
Barman can perform remote backups of multiple servers in business critical
environments and helps DBAs during the recovery phase.
# CONFIGURATION FILE LOCATIONS
The system-level Barman configuration file is located at
/etc/barman.conf
or
/etc/barman/barman.conf
and is overridden on a per-user level by
$HOME/.barman.conf
# CONFIGURATION FILE SYNTAX
The Barman configuration file is a plain `INI` file.
There is a general section called `[barman]` and a
section `[servername]` for each server you want to backup.
Rows starting with `;` are comments.
# CONFIGURATION FILE DIRECTORY
Barman supports the inclusion of multiple configuration files, through
the `configuration_files_directory` option. Included files must contain
only server specifications, not global configurations.
If the value of `configuration_files_directory` is a directory, Barman reads
all files with `.conf` extension that exist in that folder.
For example, if you set it to `/etc/barman.d`, you can
specify your PostgreSQL servers placing each section in a separate `.conf`
file inside the `/etc/barman.d` folder.
# OPTIONS
active
: When set to `true` (default), the server is in full operational state.
When set to `false`, the server can be used for diagnostics, but any
operational command such as backup execution or WAL archiving is
temporarily disabled. Setting `active=false` is a good practice
when adding a new node to Barman. Server.
archiver
: This option allows you to activate log file shipping through PostgreSQL's
`archive_command` for a server. If set to `true` (default), Barman expects
that continuous archiving for a server is in place and will activate
checks as well as management (including compression) of WAL files that
Postgres deposits in the *incoming* directory. Setting it to `false`,
will disable standard continuous archiving for a server. Global/Server.
(NOTE: this option is currently required to be enabled until Barman
natively supports physical replication slots. Setting it to `false`
will result in a disabled server).
archiver_batch_size
: This option allows you to activate batch processing of WAL files
for the `archiver` process, by setting it to a value > 0. Otherwise,
the traditional unlimited processing of the WAL queue is enabled.
When batch processing is activated, the `archive-wal` process would
limit itself to maximum `archiver_batch_size` WAL segments per single
run. Integer. Global/Server.
backup_directory
: Directory where backup data for a server will be placed. Server.
backup_method
: Configure the method barman used for backup execution.
If set to `rsync` (default), barman will execute backup using the `rsync`
command. If set to `postgres` barman will use the `pg_basebackup` command
to execute the backup. Global/Server.
backup_options
: This option allows you to control the way Barman interacts with PostgreSQL
for backups. If set to `exclusive_backup` (default), `barman backup`
executes backup operations using the standard exclusive backup approach
(technically through pg_start_backup/pg_stop_backup).
If set to `concurrent_backup`, if using PostgreSQL 9.2, 9.3, 9.4, and 9.5,
Barman requires the `pgespresso` module to be installed on the
PostgreSQL server and can be used to perform a backup from a standby server.
Starting from PostgreSQL 9.6, Barman uses the new PostgreSQL API to
perform backups from a standby server. Global/Server.
bandwidth_limit
: This option allows you to specify a maximum transfer rate in
kilobytes per second. A value of zero specifies no limit (default).
Global/Server.
barman_home
: Main data directory for Barman. Global.
barman_lock_directory
: Directory for locks. Default: `%(barman_home)s`. Global.
basebackups_directory
: Directory where base backups will be placed. Server.
basebackup_retry_sleep
: Number of seconds of wait after a failed copy, before retrying
Used during both backup and recovery operations.
Positive integer, default 30. Global/Server.
basebackup_retry_times
: Number of retries of base backup copy, after an error.
Used during both backup and recovery operations.
Positive integer, default 0. Global/Server.
check_timeout
: Maximum execution time, in seconds per server, for a barman check
command. Set to 0 to disable the timeout.
Positive integer, default 30. Global/Server.
compression
: Standard compression algorithm applied to WAL files. Possible values
are: `gzip` (requires `gzip` to be installed on the system),
`bzip2` (requires `bzip2`), `pigz` (requires `pigz`), `pygzip`
(Python's internal gzip compressor) and `pybzip2` (Python's internal
bzip2 compressor). Global/Server.
conninfo
: Connection string used by Barman to connect to the Postgres server.
This is a libpq connection string, consult the
[PostgreSQL manual][conninfo] for more information. Commonly used
keys are: host, hostaddr, port, dbname, user, password. Server.
[conninfo]: https://www.postgresql.org/docs/current/static/libpq-connect.html#LIBPQ-CONNSTRING
custom_compression_filter
: Customised compression algorithm applied to WAL files. Global/Server.
custom_decompression_filter
: Customised decompression algorithm applied to compressed WAL files;
this must match the compression algorithm. Global/Server.
description
: A human readable description of a server. Server.
errors_directory
: Directory that contains WAL files that contain an error; usually
this is related to a conflict with an existing WAL file (e.g. a WAL
file that has been archived after a streamed one).
immediate_checkpoint
: This option allows you to control the way PostgreSQL handles
checkpoint at the start of the backup.
If set to `false` (default), the I/O workload for the checkpoint
will be limited, according to the `checkpoint_completion_target`
setting on the PostgreSQL server. If set to `true`, an immediate
checkpoint will be requested, meaning that PostgreSQL will complete
the checkpoint as soon as possible. Global/Server.
incoming_wals_directory
: Directory where incoming WAL files are archived into.
Requires `archiver` to be enabled. Server.
last_backup_maximum_age
: This option identifies a time frame that must contain the latest backup.
If the latest backup is older than the time frame, barman check command
will report an error to the user.
If empty (default), latest backup is always considered valid.
Syntax for this option is: "i (DAYS | WEEKS | MONTHS)" where i is a integer
greater than zero, representing the number of days | weeks | months
of the time frame. Global/Server.
log_file
: Location of Barman's log file. Global.
log_level
: Level of logging (DEBUG, INFO, WARNING, ERROR, CRITICAL). Global.
minimum_redundancy
: Minimum number of backups to be retained. Default 0. Global/Server.
network_compression
: This option allows you to enable data compression for network
transfers.
If set to `false` (default), no compression is used.
If set to `true`, compression is enabled, reducing network usage.
Global/Server.
path_prefix
: One or more absolute paths, separated by colon, where Barman looks for
executable files. The paths specified in `path_prefix` are tried before
the ones specified in `PATH` environment variable. Global/server.
post_archive_retry_script
: Hook script launched after a WAL file is archived by maintenance.
Being this a _retry_ hook script, Barman will retry the execution of the
script until this either returns a SUCCESS (0), an ABORT_CONTINUE (62) or
an ABORT_STOP (63) code. In a post archive scenario, ABORT_STOP
has currently the same effects as ABORT_CONTINUE. Global/Server.
post_archive_script
: Hook script launched after a WAL file is archived by maintenance,
after 'post_archive_retry_script'. Global/Server.
post_backup_retry_script
: Hook script launched after a base backup.
Being this a _retry_ hook script, Barman will retry the execution of the
script until this either returns a SUCCESS (0), an ABORT_CONTINUE (62) or
an ABORT_STOP (63) code. In a post backup scenario, ABORT_STOP
has currently the same effects as ABORT_CONTINUE. Global/Server.
post_backup_script
: Hook script launched after a base backup, after 'post_backup_retry_script'.
Global/Server.
pre_archive_retry_script
: Hook script launched before a WAL file is archived by maintenance,
after 'pre_archive_script'.
Being this a _retry_ hook script, Barman will retry the execution of the
script until this either returns a SUCCESS (0), an ABORT_CONTINUE (62) or
an ABORT_STOP (63) code. Returning ABORT_STOP will propagate the failure at
a higher level and interrupt the WAL archiving operation. Global/Server.
pre_archive_script
: Hook script launched before a WAL file is archived by maintenance.
Global/Server.
pre_backup_retry_script
: Hook script launched before a base backup, after 'pre_backup_script'.
Being this a _retry_ hook script, Barman will retry the execution of the
script until this either returns a SUCCESS (0), an ABORT_CONTINUE (62) or
an ABORT_STOP (63) code. Returning ABORT_STOP will propagate the failure at
a higher level and interrupt the backup operation. Global/Server.
pre_backup_script
: Hook script launched before a base backup. Global/Server.
recovery_options
: Options for recovery operations. Currently only supports `get-wal`.
`get-wal` activates generation of a basic `restore_command` in
the resulting `recovery.conf` file that uses the `barman get-wal`
command to fetch WAL files directly from Barman's archive of WALs.
Comma separated list of values, default empty. Global/Server.
retention_policy
: Policy for retention of periodic backups and archive logs. If left empty,
retention policies are not enforced. For redundancy based retention policy
use "REDUNDANCY i" (where i is an integer > 0 and defines the number
of backups to retain). For recovery window retention policy use
"RECOVERY WINDOW OF i DAYS" or "RECOVERY WINDOW OF i WEEKS" or
"RECOVERY WINDOW OF i MONTHS" where i is a positive integer representing,
specifically, the number of days, weeks or months to retain your backups.
For more detailed information, refer to the official documentation.
Default value is empty. Global/Server.
retention_policy_mode
: Currently only "auto" is implemented. Global/Server.
reuse_backup
: This option controls incremental backup support. Global/Server.
Possible values are:
* `off`: disabled (default);
* `copy`: reuse the last available backup for a server and
create a copy of the unchanged files (reduce backup time);
* `link`: reuse the last available backup for a server and
create a hard link of the unchanged files (reduce backup time
and space). Requires operating system and file system support
for hard links.
slot_name
: Physical replication slot to be used by the `receive-wal`
command when `streaming_archiver` is set to `on`. Requires
PostgreSQL >= 9.4. Global/Server. Default: None (disabled).
streaming_archiver
: This option allows you to use the PostgreSQL's streaming protocol to
receive transaction logs from a server. If set to `on`, Barman expects
to find `pg_receivexlog` in the PATH (see `path_prefix` option) and that
streaming connection for the server is working. This activates connection
checks as well as management (including compression) of WAL files.
If set to `off` (default) barman will rely only on continuous archiving
for a server WAL archive operations, eventually terminating any running
`pg_receivexlog` for the server. Global/Server.
streaming_archiver_batch_size
: This option allows you to activate batch processing of WAL files
for the `streaming_archiver` process, by setting it to a value > 0.
Otherwise, the traditional unlimited processing of the WAL queue
is enabled. When batch processing is activated, the `archive-wal`
process would limit itself to maximum `streaming_archiver_batch_size`
WAL segments per single run. Integer. Global/Server.
streaming_archiver_name
: Identifier to be used as `application_name` by the `receive-wal` command.
Only available with `pg_receivexlog` >= 9.3. By default it is set to
`barman_receive_wal`. Global/Server.
streaming_backup_name
: Identifier to be used as `application_name` by the `pg_basebackup` command.
Only available with `pg_basebackup` >= 9.3. By default it is set to
`barman_streaming_backup`. Global/Server.
streaming_conninfo
: Connection string used by Barman to connect to the Postgres server via
streaming replication protocol. By default it is set to `conninfo`. Server.
streaming_wals_directory
: Directory where WAL files are streamed from the PostgreSQL server
to Barman. Requires `streaming_archiver` to be enabled. Server.
ssh_command
: Command used by Barman to login to the Postgres server via ssh. Server.
tablespace_bandwidth_limit
: This option allows you to specify a maximum transfer rate in
kilobytes per second, by specifying a comma separated list of
tablespaces (pairs TBNAME:BWLIMIT). A value of zero specifies no limit
(default). Global/Server.
wal_retention_policy
: Policy for retention of archive logs (WAL files). Currently only "MAIN"
is available. Global/Server.
wals_directory
: Directory which contains WAL files. Server.
# HOOK SCRIPTS
The script definition is passed to a shell and can return any exit code.
The shell environment will contain the following variables:
`BARMAN_CONFIGURATION`
: configuration file used by barman
`BARMAN_ERROR`
: error message, if any (only for the 'post' phase)
`BARMAN_PHASE`
: 'pre' or 'post'
`BARMAN_RETRY`
: `1` if it is a _retry script_ (from 1.5.0), `0` if not
`BARMAN_SERVER`
: name of the server
Backup scripts specific variables:
`BARMAN_BACKUP_DIR`
: backup destination directory
`BARMAN_BACKUP_ID`
: ID of the backup
`BARMAN_PREVIOUS_ID`
: ID of the previous backup (if present)
`BARMAN_STATUS`
: status of the backup
`BARMAN_VERSION`
: version of Barman
Archive scripts specific variables:
`BARMAN_SEGMENT`
: name of the WAL file
`BARMAN_FILE`
: full path of the WAL file
`BARMAN_SIZE`
: size of the WAL file
`BARMAN_TIMESTAMP`
: WAL file timestamp
`BARMAN_COMPRESSION`
: type of compression used for the WAL file
Only in case of retry hook scripts, the exit code of the script
is checked by Barman. Output of hook scripts is simply written
in the log file.
# EXAMPLE
Here is an example of configuration file:
```
[barman]
; Main directory
barman_home = /var/lib/barman
; System user
barman_user = barman
; Log location
log_file = /var/log/barman/barman.log
; Default compression level
;compression = gzip
; Incremental backup
reuse_backup = link
; 'main' PostgreSQL Server configuration
[main]
; Human readable description
description = "Main PostgreSQL Database"
; SSH options
ssh_command = ssh postgres@pg
; PostgreSQL connection string
conninfo = host=pg user=postgres
; PostgreSQL streaming connection string
streaming_conninfo = host=pg user=postgres
; Minimum number of required backups (redundancy)
minimum_redundancy = 1
; Retention policy (based on redundancy)
retention_policy = REDUNDANCY 2
```
# SEE ALSO
`barman` (1).
# AUTHORS
In alphabetical order:
* Gabriele Bartolini (project leader)
* Jonathan Battiato (QA/testing)
* Stefano Bianucci (developer, intern from University of Florence)
* Giuseppe Broccolo (QA/testing)
* Giulio Calacoci (developer)
* Francesco Canovai (QA/testing)
* Leonardo Cecchi (developer)
* Gianni Ciolli (QA/testing)
* Britt Cole (documentation)
* Marco Nenciarini (lead developer)
* Rubens Souza (QA/testing)
Past contributors:
* Carlo Ascani
# RESOURCES
* Homepage:
* Documentation:
* Professional support:
# COPYING
Barman is the exclusive property of 2ndQuadrant Italia
and its code is distributed under GNU General Public License v3.
Copyright (C) 2011-2016 2ndQuadrant Italia Srl - http://www.2ndQuadrant.it/.
barman-2.1/doc/barman.d/ 0000775 0001750 0001750 00000000000 13033432341 015435 5 ustar gcalacoci gcalacoci barman-2.1/doc/barman.d/ssh-server.conf-template 0000664 0001750 0001750 00000002574 13015546105 022232 0 ustar gcalacoci gcalacoci ; Barman, Backup and Recovery Manager for PostgreSQL
; http://www.pgbarman.org/ - http://www.2ndQuadrant.com/
;
; Template configuration file for a server using
; SSH connections and rsync for copy.
;
[ssh]
; Human readable description
description = "Example of PostgreSQL Database (via SSH)"
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; SSH options (mandatory)
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
ssh_command = ssh postgres@pg
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; PostgreSQL connection string (mandatory)
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
conninfo = host=pg user=barman dbname=postgres
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; Backup settings (via rsync over SSH)
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
backup_method = rsync
; Incremental backup support: possible values are None (default), link or copy
;reuse_backup = link
; Identify the standard behavior for backup operations: possible values are
; exclusive_backup (default), concurrent_backup
;backup_options = exclusive_backup
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; Continuous WAL archiving (via 'archive_command')
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
archiver = on
;archiver_batch_size = 50
; PATH setting for this server
;path_prefix = "/usr/pgsql-9.6/bin"
barman-2.1/doc/barman.d/streaming-server.conf-template 0000664 0001750 0001750 00000002701 13015546105 023416 0 ustar gcalacoci gcalacoci ; Barman, Backup and Recovery Manager for PostgreSQL
; http://www.pgbarman.org/ - http://www.2ndQuadrant.com/
;
; Template configuration file for a server using
; only streaming replication protocol
;
[streaming]
; Human readable description
description = "Example of PostgreSQL Database (Streaming-Only)"
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; PostgreSQL connection string (mandatory)
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
conninfo = host=pg user=barman dbname=postgres
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; PostgreSQL streaming connection string
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; To be used by pg_basebackup for backup and pg_receivexlog for WAL streaming
; NOTE: streaming_barman is a regular user with REPLICATION privilege
streaming_conninfo = host=pg user=streaming_barman
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; Backup settings (via pg_basebackup)
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
backup_method = postgres
;streaming_backup_name = barman_streaming_backup
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; WAL streaming settings (via pg_receivexlog)
; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
streaming_archiver = on
slot_name = barman
;streaming_archiver_name = barman_receive_wal
;streaming_archiver_batch_size = 50
; PATH setting for this server
;path_prefix = "/usr/pgsql-9.6/bin"
barman-2.1/doc/Makefile 0000664 0001750 0001750 00000000670 13015546105 015422 0 ustar gcalacoci gcalacoci MANPAGES=barman.1 barman.5
DOCS=
SUBDIRS=manual
.PHONY: all clean help subdirs $(SUBDIRS)
all: $(MANPAGES) $(DOCS) $(SUBDIRS)
barman.1: barman.1.md
pandoc -s -t man -o $@ $<
barman.5: barman.5.md
pandoc -s -t man -o $@ $<
clean:
rm -f $(MANPAGES) $(DOCS)
for dir in $(SUBDIRS); do \
$(MAKE) -C $$dir clean; \
done
help:
@echo "Usage:"
@echo " $$ make"
subdirs: $(SUBDIRS)
$(SUBDIRS):
$(MAKE) -C $@
barman-2.1/doc/barman.1.md 0000664 0001750 0001750 00000033367 13033425212 015707 0 ustar gcalacoci gcalacoci % BARMAN(1) Barman User manuals | Version 2.1
% 2ndQuadrant Italy
% January 5, 2017
# NAME
barman - Backup and Recovery Manager for PostgreSQL
# SYNOPSIS
barman [*OPTIONS*] *COMMAND*
# DESCRIPTION
Barman is an administration tool for disaster recovery of PostgreSQL
servers written in Python and maintained by 2ndQuadrant.
Barman can perform remote backups of multiple servers in business critical
environments and helps DBAs during the recovery phase.
# OPTIONS
-v, --version
: Show program version number and exit.
-q, --quiet
: Do not output anything. Useful for cron scripts.
-h, --help
: Show a help message and exit.
-c *CONFIG*, --config *CONFIG*
: Use the specified configuration file.
# COMMANDS
Important: every command has a help option
archive-wal *SERVER_NAME*
: Get any incoming xlog file (both through standard `archive_command`
and streaming replication, where applicable) and moves them in the
WAL archive for that server. If necessary, apply compression when
requested by the user.
cron
: Perform maintenance tasks, such as enforcing retention policies or
WAL files management.
list-server
: Show all the configured servers, and their descriptions.
show-server *SERVER_NAME*
: Show information about `SERVER_NAME`, including: `conninfo`,
`backup_directory`, `wals_directory` and many more.
Specify `all` as `SERVER_NAME` to show information about all
the configured servers.
status *SERVER_NAME*
: Show information about the status of a server, including: number of
available backups, `archive_command`, `archive_status` and many more.
For example:
```
Server quagmire:
Description: The Giggity database
Passive node: False
PostgreSQL version: 9.3.9
pgespresso extension: Not available
PostgreSQL Data directory: /srv/postgresql/9.3/data
PostgreSQL 'archive_command' setting: rsync -a %p barman@backup:/var/lib/barman/quagmire/incoming
Last archived WAL: 0000000100003103000000AD
Current WAL segment: 0000000100003103000000AE
Retention policies: enforced (mode: auto, retention: REDUNDANCY 2, WAL retention: MAIN)
No. of available backups: 2
First available backup: 20150908T003001
Last available backup: 20150909T003001
Minimum redundancy requirements: satisfied (2/1)
```
check *SERVER_NAME*
: Show diagnostic information about `SERVER_NAME`, including:
Ssh connection check, PostgreSQL version, configuration and backup
directories, archiving process, streaming process, replication slots, etc.
Specify `all` as `SERVER_NAME` to show diagnostic information
about all the configured servers.
--nagios
: Nagios plugin compatible output
diagnose
: Collect diagnostic information about the server where barman is installed
and all the configured servers, including: global configuration, SSH version,
Python version, `rsync` version, as well as current configuration and status
of all servers.
backup *SERVER_NAME*
: Perform a backup of `SERVER_NAME` using parameters specified in the
configuration file. Specify `all` as `SERVER_NAME` to perform a backup
of all the configured servers.
--immediate-checkpoint
: forces the initial checkpoint to be done as quickly as possible.
Overrides value of the parameter `immediate_checkpoint`, if present
in the configuration file.
--no-immediate-checkpoint
: forces to wait for the checkpoint.
Overrides value of the parameter `immediate_checkpoint`, if present
in the configuration file.
--reuse-backup [INCREMENTAL_TYPE]
: Overrides `reuse_backup` option behaviour. Possible values for
`INCREMENTAL_TYPE` are:
- *off*: do not reuse the last available backup;
- *copy*: reuse the last available backup for a server and
create a copy of the unchanged files (reduce backup time);
- *link*: reuse the last available backup for a server and
create a hard link of the unchanged files (reduce backup time
and space);
`link` is the default target if `--reuse-backup` is used and
`INCREMENTAL_TYPE` is not explicited.
--retry-times
: Number of retries of base backup copy, after an error.
Used during both backup and recovery operations.
Overrides value of the parameter `basebackup_retry_times`,
if present in the configuration file.
--no-retry
: Same as `--retry-times 0`
--retry-sleep
: Number of seconds of wait after a failed copy, before retrying.
Used during both backup and recovery operations.
Overrides value of the parameter `basebackup_retry_sleep`,
if present in the configuration file.
list-backup *SERVER_NAME*
: Show available backups for `SERVER_NAME`. This command is useful to
retrieve a backup ID. For example:
```
servername 20111104T102647 - Fri Nov 4 10:26:48 2011 - Size: 17.0 MiB - WAL Size: 100 B
```
In this case, *20111104T102647* is the backup ID.
show-backup *SERVER_NAME* *BACKUP_ID*
: Show detailed information about a particular backup, identified by
the server name and the backup ID. See the [Backup ID shortcuts](#shortcuts)
section below for available shortcuts. For example:
```
Backup 20150828T130001:
Server Name : quagmire
Status : DONE
PostgreSQL Version : 90402
PGDATA directory : /srv/postgresql/9.4/main/data
Base backup information:
Disk usage : 12.4 TiB (12.4 TiB with WALs)
Incremental size : 4.9 TiB (-60.02%)
Timeline : 1
Begin WAL : 0000000100000CFD000000AD
End WAL : 0000000100000D0D00000008
WAL number : 3932
WAL compression ratio: 79.51%
Begin time : 2015-08-28 13:00:01.633925+00:00
End time : 2015-08-29 10:27:06.522846+00:00
Begin Offset : 1575048
End Offset : 13853016
Begin XLOG : CFD/AD180888
End XLOG : D0D/8D36158
WAL information:
No of files : 35039
Disk usage : 121.5 GiB
WAL rate : 275.50/hour
Compression ratio : 77.81%
Last available : 0000000100000D95000000E7
Catalog information:
Retention Policy : not enforced
Previous Backup : 20150821T130001
Next Backup : - (this is the latest base backup)
```
list-files *\[OPTIONS\]* *SERVER_NAME* *BACKUP_ID*
: List all the files in a particular backup, identified by the server
name and the backup ID. See the [Backup ID shortcuts](#shortcuts) section below
for available shortcuts.
--target *TARGET_TYPE*
: Possible values for TARGET_TYPE are:
- *data*: lists just the data files;
- *standalone*: lists the base backup files, including required
WAL files;
- *wal*: lists all the WAL files between the start of the base
backup and the end of the log / the start of the following base
backup (depending on whether the specified base backup is the most
recent one available);
- *full*: same as data + wal.
The default value is `standalone`.
rebuild-xlogdb *SERVER_NAME*
: Perform a rebuild of the WAL file metadata for `SERVER_NAME`
(or every server, using the `all` shortcut) guessing it from
the disk content. The metadata of the WAL archive is contained
in the `xlog.db` file, and every Barman server has its own copy.
recover *\[OPTIONS\]* *SERVER_NAME* *BACKUP_ID* *DESTINATION_DIRECTORY*
: Recover a backup in a given directory (local or remote, depending
on the `--remote-ssh-command` option settings).
See the [Backup ID shortcuts](#shortcuts) section below for available shortcuts.
--target-tli *TARGET_TLI*
: Recover the specified timeline.
--target-time *TARGET_TIME*
: Recover to the specified time.
You can use any valid unambiguous representation
(e.g: "YYYY-MM-DD HH:MM:SS.mmm").
--target-xid *TARGET_XID*
: Recover to the specified transaction ID.
--target-name *TARGET_NAME*
: Recover to the named restore point previously created with
the `pg_create_restore_point(name)` (for PostgreSQL 9.1 and above users).
--exclusive
: Set target xid to be non inclusive.
--tablespace *NAME:LOCATION*
: Specify tablespace relocation rule.
--remote-ssh-command *SSH_COMMAND*
: This options activates remote recovery, by specifying the
secure shell command to be launched on a remote host.
This is the equivalent of the "ssh_command" server option
in the configuration file for remote recovery.
Example: 'ssh postgres@db2'.
--retry-times *RETRY_TIMES*
: Number of retries of data copy during base backup after
an error. Overrides value of the parameter `basebackup_retry_times`,
if present in the configuration file.
--no-retry
: Same as `--retry-times 0`
--retry-sleep
: Number of seconds of wait after a failed copy, before retrying.
Overrides value of the parameter `basebackup_retry_sleep`,
if present in the configuration file.
get-wal *\[OPTIONS\]* *SERVER_NAME* *WAL_ID*
: Retrieve a WAL file from the `xlog` archive of a given server.
By default, the requested WAL file, if found, is returned as
uncompressed content to `STDOUT`. The following options allow
users to change this behaviour:
-o *OUTPUT_DIRECTORY*
: destination directory where the `get-wal` will deposit the requested WAL
-z
: output will be compressed using gzip
-j
: output will be compressed using bzip2
-p *SIZE*
: peek from the WAL archive up to *SIZE* WAL files, starting
from the requested one. 'SIZE' must be an integer >= 1.
When invoked with this option, get-wal returns a
list of zero to 'SIZE' WAL segment names, one per row.
switch-xlog *SERVER_NAME*
: Execute pg_switch_xlog() on the target server
--force
: Forces the switch by executing CHECKPOINT before pg_switch_xlog().
*IMPORTANT:* executing a CHECKPOINT might increase I/O load on
a PostgreSQL server. Use this option with care.
--archive
: Wait for one xlog file to be archived.
If after a defined amount of time (default: 30 seconds) no xlog
file is archived, Barman will teminate with failure exit code.
--archive-timeout *TIMEOUT*
: Specifies the amount of time in seconds (default: 30 seconds)
the archiver will wait for a new xlog file to be archived
before timing out.
receive-wal *SERVER_NAME*
: Start the stream of transaction logs for a server.
The process relies on `pg_receivexlog` to receive WAL files
from the PostgreSQL servers through the streaming protocol.
--stop
: stop the receive-wal process for the server
--reset
: reset the status of receive-wal, restarting the streaming
from the current WAL file of the server
--create-slot
: create the physical replication slot configured with the
`slot_name` configuration parameter
--drop-slot
: drop the physical replication slot configured with the
`slot_name` configuration parameter
delete *SERVER_NAME* *BACKUP_ID*
: Delete the specified backup. [Backup ID shortcuts](#shortcuts)
section below for available shortcuts.
replication-status *\[OPTIONS\]* *SERVER_NAME*
: Shows live information and status of any streaming client attached
to the given server (or servers). Default behaviour can be changed
through the following options:
--minimal
: machine readable output (default: False)
--target *TARGET_TYPE*
: Possible values for TARGET_TYPE are:
- *hot-standby*: lists only hot standby servers
- *wal-streamer*: lists only WAL streaming clients, such as
pg_receivexlog
- *all*: any streaming client (default)
# BACKUP ID SHORTCUTS {#shortcuts}
Rather than using the timestamp backup ID, you can use any of the
following shortcuts/aliases to identity a backup for a given server:
first
: Oldest available backup for that server, in chronological order.
last
: Latest available backup for that server, in chronological order.
latest
: same ast *last*.
oldest
: same ast *first*.
# EXIT STATUS
0
: Success
Not zero
: Failure
# SEE ALSO
`barman` (5).
# BUGS
Barman has been extensively tested, and is currently being used in several
production environments. However, we cannot exclude the presence of bugs.
Any bug can be reported via the Sourceforge bug tracker. Along the bug submission,
users can provide developers with diagnostics information obtained through
the `barman diagnose` command.
# AUTHORS
In alphabetical order:
* Gabriele Bartolini (project leader)
* Jonathan Battiato (QA/testing)
* Stefano Bianucci (developer, intern from University of Florence)
* Giuseppe Broccolo (QA/testing)
* Giulio Calacoci (developer)
* Francesco Canovai (QA/testing)
* Leonardo Cecchi (developer)
* Gianni Ciolli (QA/testing)
* Britt Cole (documentation)
* Marco Nenciarini (lead developer)
* Rubens Souza (QA/testing)
Past contributors:
* Carlo Ascani
# RESOURCES
* Homepage:
* Documentation:
* Professional support:
# COPYING
Barman is the property of 2ndQuadrant Italia
and its code is distributed under GNU General Public License v3.
Copyright (C) 2011-2016 2ndQuadrant Italia Srl - .
barman-2.1/doc/manual/ 0000775 0001750 0001750 00000000000 13033432341 015230 5 ustar gcalacoci gcalacoci barman-2.1/doc/manual/00-head.en.md 0000664 0001750 0001750 00000001416 13033425212 017272 0 ustar gcalacoci gcalacoci % Barman Manual
% 2ndQuadrant Italia
% January 5, 2017 (v2.1)
**Barman** (Backup and Recovery Manager) is an open-source administration tool for disaster recovery of PostgreSQL servers written in Python. It allows your organisation to perform remote backups of multiple servers in business critical environments to reduce risk and help DBAs during the recovery phase.
[Barman] [11] is distributed under GNU GPL 3 and maintained by [2ndQuadrant] [13], a platinum sponsor of the [PostgreSQL project] [31].
> **IMPORTANT:** \newline
> This manual assumes that you are familiar with theoretical disaster
> recovery concepts, and that you have a grasp of PostgreSQL fundamentals in
> terms of physical backup and disaster recovery. See section _"Before you start"_ below for details.
barman-2.1/doc/manual/43-backup-commands.en.md 0000664 0001750 0001750 00000012522 13015546106 021452 0 ustar gcalacoci gcalacoci \newpage
# Backup commands
Backup commands are those that works directly on backups already existing in
Barman's backup catalog.
> **NOTE:**
> Remember a backup ID can be retrieved with `barman list-backup
> `
## Backup ID shortcuts
Barman allows you to use special keywords to identify a specific backup:
* `last/latest`: identifies the newest backup in the catalog
* `first/oldest`: identifies the oldest backup in the catalog
Using those keywords with Barman commands allows you to execute actions
without knowing the exact ID of a backup for a server.
For example we can issue:
``` bash
barman delete oldest
```
to remove the oldest backup available in the catalog and reclaim disk space.
## `delete`
You can delete a given backup with:
``` bash
barman delete
```
The `delete` command accepts any [shortcut](#shortcuts) to identify backups.
## `list-files`
You can list the files (base backup and required WAL files) for a
given backup with:
``` bash
barman list-files [--target TARGET_TYPE]
```
With the `--target TARGET_TYPE` option, it is possible to choose the
content of the list for a given backup.
Possible values for `TARGET_TYPE` are:
- `data`: lists the data files
- `standalone`: lists the base backup files, including required WAL
files
- `wal`: lists all WAL files from the beginning of the base backup to
the start of the following one (or until the end of the log)
- `full`: same as `data` + `wal`
The default value for `TARGET_TYPE` is `standalone`.
> **IMPORTANT:**
> The `list-files` command facilitates interaction with external
> tools, and can therefore be extremely useful to integrate
> Barman into your archiving procedures.
## `recover`
The `recover` command is used to recover a whole server after
a backup is executed using the `backup` command.
This is achieved issuing a command like the following:
```bash
barman@backup$ barman recover /path/to/recover/dir
```
At the end of the execution of the recovery, the selected backup is recovered
locally and the destination path contains a data directory ready to be used
to start a PostgreSQL instance.
> **IMPORTANT:**
> Running this command as user `barman`, it will become the database superuser.
The specific ID of a backup can be retrieved using the [list-backup](#list-backup)
command.
> **IMPORTANT:**
> Barman does not currently keep track of symbolic links inside PGDATA
> (except for tablespaces inside pg_tblspc). We encourage
> system administrators to keep track of symbolic links and to add them
> to the disaster recovery plans/procedures in case they need to be restored
> in their original location.
The recovery command has several options that modify the command behavior.
### Remote recovery
Add the `--remote-ssh-command ` option to the invocation
of the recovery command. Doing this will allow Barman to execute
the copy on a remote server, using the provided command to connect
to the remote host.
> **NOTE:**
> It is advisable to use the `postgres` user to perform
> the recovery on the remote host.
Known limitations of the remote recovery are:
* Barman requires at least 4GB of free space in the system temporary directory
unless the [`get-wal`](#get-wal) command is specified
in the `recovery_option` parameter in the Barman configuration.
* The SSH connection between Barman and the remote host **must** use the
public key exchange authentication method
* The remote user **must** be able to create the directory structure
of the backup in the destination directory.
* There must be enough free space on the remote server
to contain the base backup and the WAL files needed for recovery.
### Tablespace remapping
Barman is able to automatically remap one or more tablespaces using
the recover command with the --tablespace option.
The option accepts a pair of values as arguments using the
`NAME:DIRECTORY` format:
* `NAME` is the identifier of the tablespace
* `DIRECTORY` is the new destination path for the tablespace
If the destination directory does not exists,
Barman will try to create it (assuming you have the required permissions).
### Point in time recovery
Barman wraps PostgreSQL's Point-in-Time Recovery (PITR),
allowing you to specify a recovery target, either as a timestamp,
as a restore label, or as a transaction ID.
The recovery target can be specified using one of
three mutually exclusive options:
* --target-time TARGET_TIME: to specify a timestamp
* --target-xid TARGET_XID: to specify a transaction ID
* --target-name TARGET_NAME: to specify a named restore point
previously created with the pg_create_restore_point(name)
function[^TARGET_NAME]
[^TARGET_NAME]:
Only available on PostgreSQL 9.1 and above
You can use the --exclusive option to specify whether to stop immediately
before or immediately after the recovery target.
Barman allows you to specify a target timeline for recovery,
using the `target-tli` option. The notion of timeline goes beyond the scope of
this document; you can find more details in the PostgreSQL documentation,
as mentioned in the _"Before you start"_ section.
## `show-backup`
You can retrieve all the available information for a particular backup of
a given server with:
``` bash
barman show-backup
```
The `show-backup` command accepts any [shortcut](#shortcuts) to identify backups.
barman-2.1/doc/manual/10-design.en.md 0000664 0001750 0001750 00000023170 13015546106 017652 0 ustar gcalacoci gcalacoci \newpage
# Design and architecture
## Where to install Barman
One of the foundations of Barman is the ability to operate remotely from the database server, via the network.
Theoretically, you could have your Barman server located in a data centre in another part of the world, thousands of miles away from your PostgreSQL server.
Realistically, you do not want your Barman server to be too far from your PostgreSQL server, so that both backup and recovery times are kept under control.
Even though there is no _"one size fits all"_ way to setup Barman, there are a couple of recommendations that we suggest you abide by, in particular:
- Install Barman on a dedicated server
- Do not share the same storage with your PostgreSQL server
- Integrate Barman with your monitoring infrastructure [^nagios]
- Test everything before you deploy it to production
[^nagios]: Integration with Nagios/Icinga is straightforward thanks to the `barman check --nagios` command, one of the most important features of Barman and a true lifesaver.
A reasonable way to start modelling your disaster recovery architecture is to:
- design a couple of possibile architectures in respect to PostgreSQL and Barman, such as:
1. same data centre
2. different data centre in the same metropolitan area
3. different data centre
- elaborate the pros and the cons of each hypothesis
- evaluate the single points of failure (SPOF) of your system, with cost-benefit analysis
- make your decision and implement the initial solution
Having said this, a very common setup for Barman is to be installed in the same data centre where your PostgreSQL servers are. In this case, the single point of failure is the data centre. Fortunately, the impact of such a SPOF can be alleviated thanks to a feature called _hook scripts_. Indeed, backups of Barman can be exported on different media, such as _tape_ via `tar`, or locations, like an _S3 bucket_ in the Amazon cloud.
Remember that no decision is forever. You can start this way and adapt over time to the solution that suits you best. However, try and keep it simple to start with.
## One Barman, many PostgreSQL servers
Another relevant feature that was first introduced by Barman is support for multiple servers. Barman can store backup data coming from multiple PostgreSQL instances, even with different versions, in a centralised way. [^recver]
[^recver]: The same [requirements for PostgreSQL's PITR][requirements_recovery] apply for recovery, as detailed in the section _"Requirements for recovery"_.
As a result, you can model complex disaster recovery architectures, forming a "star schema", where PostgreSQL servers rotate around a central Barman server.
Every architecture makes sense in its own way. Choose the one that resonates with you, and most importantly, the one you trust, based on real experimentation and testing.
From this point forward, for the sake of simplicity, this guide will assume a basic architecture:
- one PostgreSQL instance (with host name `pg`)
- one backup server with Barman (with host name `backup`)
## Streaming backup vs rsync/SSH
Traditionally, Barman has always operated remotely via SSH, taking advantage of `rsync` for physical backup operations. Version 2.0 introduces native support for PostgreSQL's streaming replication protocol for backup operations, via `pg_basebackup`. [^fmatrix]
[^fmatrix]: Check in the "Feature matrix" which PostgreSQL versions support streaming replication backups with Barman.
Choosing one of these two methods is a decision you will need to make.
On a general basis, starting from Barman 2.0, backup over streaming replication is the recommended setup for PostgreSQL 9.4 or higher. Moreover, if you do not make use of tablespaces, backup over streaming can be used starting from PostgreSQL 9.2.
> **IMPORTANT:** \newline
> Because Barman transparently makes use of `pg_basebackup`, features such as incremental backup, deduplication, and network compression are currently not available. In this case, bandwidth limitation has some restrictions - compared to the traditional method via `rsync`.
Traditional backup via `rsync`/SSH is available for all versions of PostgreSQL starting from 8.3, and it is recommended in all cases where `pg_basebackup` limitations occur (for example, a very large database that can benefit from incremental backup and deduplication).
The reason why we recommend streaming backup is that, based on our experience, it is easier to setup than the traditional one. Also, streaming backup allows you to backup a PostgreSQL server on Windows[^windows], and makes life easier when working with Docker.
[^windows]: Backup of a PostgreSQL server on Windows is possible, but it is still experimental because it is not yet part of our continuous integration system. See section _"How to setup a Windows based server"_ for details.
## Standard archiving, WAL streaming ... or both
PostgreSQL's Point-In-Time-Recovery requires that transactional logs, also known as _xlog_ or WAL files, are stored alongside of base backups.
Traditionally, Barman has supported standard WAL file shipping through PostgreSQL's `archive_command` (usually via `rsync`/SSH). With this method, WAL files are archived only when PostgreSQL _switches_ to a new WAL file. To keep it simple, this normally happens every 16MB worth of data changes.
Barman 1.6.0 introduces streaming of WAL files for PostgreSQL servers 9.2 or higher, as an additional method for transactional log archiving, through `pg_receivexlog`. WAL streaming is able to reduce the risk of data loss, bringing RPO down to _near zero_ values.
Barman 2.0 introduces support for replication slots with PostgreSQL servers 9.4 or above, therefore allowing WAL streaming-only configurations. Moreover, you can now add Barman as a synchronous WAL receiver in your PostgreSQL 9.5 (or higher) cluster, and achieve **zero data loss** (RPO=0).
In some cases you have no choice and you are forced to use traditional archiving. In others, you can choose whether to use both or just WAL streaming.
Unless you have strong reasons not to do it, we recommend to use both channels, for maximum reliability and robustness.
## Two typical scenarios for backups
In order to make life easier for you, below we summarise the two most typical scenarios for a given PostgreSQL server in Barman.
Bear in mind that this is a decision that you must make for every single server that you decide to back up with Barman. This means that you can have heterogeneous setups within the same installation.
As mentioned before, we will only worry about the PostgreSQL server (`pg`) and the Barman server (`backup`). However, in real life, your architecture will most likely contain other technologies such as repmgr, pgBouncer, Nagios/Icinga, and so on.
### Scenario 1: Backup via streaming protocol
If you are using PostgreSQL 9.4 or higher, and your database falls under a general use case scenario, you will likely end up deciding on a streaming backup installation - see figure \ref{scenario1-design} below.
{ width=80% }
In this scenario, you will need to configure:
1. a standard connection to PostgreSQL, for management, coordination, and monitoring purposes
2. a streaming replication connection that will be used by both `pg_basebackup` (for base backup operations) and `pg_receivexlog` (for WAL streaming)
This setup, in Barman's terminology, is known as **streaming-only** setup, as it does not require any SSH connection for backup and archiving operations. This is particularly suitable and extremely practical for Docker environments.
However, as mentioned before, you can configure standard archiving as well and implement a more robust architecture - see figure \ref{scenario1b-design} below.
{ width=80% }
This alternate approach requires:
- an additional SSH connection that allows the `postgres` user on the PostgreSQL server to connect as `barman` user on the Barman server
- the `archive_command` in PostgreSQL be configured to ship WAL files to Barman
This architecture is available also to PostgreSQL 9.2/9.3 users that do not use tablespaces.
### Scenario 2: Backup via `rsync`/SSH
The _traditional_ setup of `rsync` over SSH is the only available option for:
- PostgreSQL servers version 8.3, 8.4, 9.0 or 9.1
- PostgreSQL servers version 9.2 or 9.3 that are using tablespaces
- incremental backup and deduplication
- network compression during backups
- finer control of bandwidth usage, including on a tablespace basis
{ width=80% }
In this scenario, you will need to configure:
1. a standard connection to PostgreSQL for management, coordination, and monitoring purposes
2. an SSH connection for base backup operations to be used by `rsync` that allows the `barman` user on the Barman server to connect as `postgres` user on the PostgreSQL server
3. an SSH connection for WAL archiving to be used by the `archive_command` in PostgreSQL and that allows the `postgres` user on the PostgreSQL server to connect as `barman` user on the Barman server
Starting from PostgreSQL 9.2, you can add a streaming replication connection that is used for WAL streaming and significantly reduce RPO. This more robust implementation is depicted in figure \ref{scenario2b-design}.
{ width=80% }
barman-2.1/doc/manual/25-streaming_backup.en.md 0000664 0001750 0001750 00000002514 13015546106 021724 0 ustar gcalacoci gcalacoci ## Streaming backup
Barman can backup a PostgreSQL server using the streaming connection,
relying on `pg_basebackup`, a utility that has been available from
PostgreSQL 9.1.
> **IMPORTANT:** Barman requires that `pg_basebackup` is installed in
> the same server. For PostgreSQL 9.2 servers, you need the
> `pg_basebackup` of version 9.2 installed alongside with Barman. For
> PostgreSQL 9.3 and above, it is recommented to install the last
> available version of `pg_basebackup`, as it is back compatible. You
> can even install multiple versions of `pg_basebackup` on the Barman
> server and properly point to the specific version for a server,
> using the `path_prefix` option in the configuration file.
To successfully backup your server with the streaming connection, you
need to use `postgres` as your backup method:
``` ini
backup_method = postgres
```
> **IMPORTANT:** keep in mind that if the WAL archiving is not
> currently configured, you will not be able to start a backup.
To check if the server configuration is valid you can use the `barman
check` command:
``` bash
barman@backup$ barman check pg
```
To start a backup you can use the `barman backup` command:
``` bash
barman@backup$ barman backup pg
```
> **IMPORTANT:** `pg_basebackup` 9.4 or higher is required for
> tablespace support if you use the `postgres` backup method.
barman-2.1/doc/manual/42-server-commands.en.md 0000664 0001750 0001750 00000017671 13025741466 021533 0 ustar gcalacoci gcalacoci \newpage
# Server commands
As we said in the previous section, server commands work directly on
a PostgreSQL server or on its area in Barman, and are useful to check
its status, perform maintainance operations, take backups, and
manage the WAL archive.
## `archive_wal`
The `archive_wal` command execute maintainance operations on WAL files
for a given server. This operations include processing of the WAL
files received from the streaming connection or from the
`archive_command` or both.
> **IMPORTANT:**
> The `archive_wal` command, even if it can be directly invoked, is
> designed to be started from the `cron` general command.
## `backup`
The `backup` command takes a full backup (_base backup_) of a given
server. It has several options that let you override the corresponding
configuration parameter for the new backup. For more information,
consult the manual page.
You can perform a full backup for a given server with:
``` bash
barman backup
```
> **TIP:**
> You can use `barman backup all` to sequentially backup all your
> configured servers.
## `check`
You can check the connection to a given server and the
configuration coherence with the `check` command:
``` bash
barman check
```
> **TIP:**
> You can use `barman check all` to check all your configured servers.
> **IMPORTANT:**
> The `check` command is probably the most critical feature that
> Barman implements. We recommend to integrate it with your alerting
> and monitoring infrastructure. The `--nagios` option allows you
> to easily create a plugin for Nagios/Icinga.
## `get-wal`
Barman allows users to request any _xlog_ file from its WAL archive
through the `get-wal` command:
``` bash
barman get-wal [-o OUTPUT_DIRECTORY] [-j|-x]
```
If the requested WAL file is found in the server archive, the
uncompressed content will be returned to `STDOUT`, unless otherwise
specified.
The following options are available for the `get-wal` command:
- `-o` allows users to specify a destination directory where Barman
will deposit the requested WAL file
- `-j` will compress the output using `bzip2` algorithm
- `-x` will compress the output using `gzip` algorithm
- `-p SIZE` peeks from the archive up to WAL files, starting from
the requested file
It is possible to use `get-wal` during a recovery operation,
transforming the Barman server into a _WAL hub_ for your servers. This
can be automatically achieved by adding the `get-wal` value to the
`recovery_options` global/server configuration option:
``` ini
recovery_options = 'get-wal'
```
`recovery_options` is a global/server option that accepts a list of
comma separated values. If the keyword `get-wal` is present during a
recovery operation, Barman will prepare the `recovery.conf` file by
setting the `restore_command` so that `barman get-wal` is used to
fetch the required WAL files.
This is an example of a `restore_command` for a local recovery:
``` ini
restore_command = 'sudo -u barman barman get-wal SERVER %f > %p'
```
Please note that the `get-wal` command should always be invoked as
`barman` user, and that it requires the correct permission to
read the WAL files from the catalog. This is the reason why we are
using `sudo -u barman` in the example.
Setting `recovery_options` to `get-wal` for a remote recovery will instead
generate a `restore_command` using the `barman-wal-restore` script.
`barman-wal-restore` is a more resilient shell script which manages SSH
connection errors.
This script has many useful options such as the automatic compression and
decompression of the WAL files and the *peek* feature, which allows you
to retrieve the next WAL files while PostgreSQL is applying one of them. It is
an excellent way to optimise the bandwidth usage between PostgreSQL and
Barman.
`barman-wal-restore` is available in the `barman-cli` project or package.
This is an example of a `restore_command` for a remote recovery:
``` ini
restore_command = 'barman-wal-restore -U barman backup SERVER %f %p'
```
Since it uses SSH to communicate with the Barman server, SSH key authentication
is required for the `postgres` user to login as `barman` on the backup server.
> **IMPORTANT:**
> Even though `recovery_options` aims to automate the process, using
> the `get-wal` facility requires manual intervention and proper
> testing.
## `list-backup`
You can list the catalog of available backups for a given server
with:
``` bash
barman list-backup
```
> **TIP:** You can request a full list of the backups of all servers
> using `all` as the server name.
To have a machine-readable output you can use the `--minimal` option.
## `rebuild-xlogdb`
At any time, you can regenerate the content of the WAL archive for a
specific server (or every server, using the `all` shortcut). The WAL
archive is contained in the `xlog.db` file and every server managed by
Barman has its own copy.
The `xlog.db` file can be rebuilt with the `rebuild-xlogdb`
command. This will scan all the archived WAL files and regenerate the
metadata for the archive.
For example:
``` bash
barman rebuild-xlogdb
```
## `receive-wal`
This command manages the `receive-wal` process, which uses the
streaming protocol to receive WAL files from the PostgreSQL streaming
connection.
### receive-wal process management
If the command is run without options, a `receive-wal` process will
be started. This command is based on the `pg_receivexlog` PostgreSQL
command.
``` bash
barman receive-wal
```
If the command is run with the `--stop` option, the currently running
`receive-wal` process will be stopped.
The `receive-wal` process uses a status file to track last written
record of the transaction log. When the status file needs to be
cleaned, the `--reset` option can be used.
> **IMPORTANT:** If you are not using replication slots, you rely
> on the value of `wal_keep_segments`. Be aware that under high peeks
> of workload on the database, the `receive-wal` process
> might fall behind and go out of sync. As a precautionary measure,
> Barman currently requires that users manually execute the command with the
> `--reset` option, to avoid making wrong assumptions.
### Replication slot management
The `receive-wal` process is also useful to create or drop the
replication slot needed by Barman for its WAL archiving procedure.
With the `--create-slot` option, the replication slot named after the
`slot_name` configuration option will be created on the PostgreSQL
server.
With the `--drop-slot`, the previous replication slot will be deleted.
## `replication-status`
The `replication-status` command reports the status of any streaming
client currently attached to the PostgreSQL server, including the
`receive-wal` process of your Barman server (if configured).
You can execute the command as follows:
``` bash
barman replication-status
```
> **TIP:** You can request a full status report of the replica
> for all your servers using `all` as the server name.
To have a machine-readable output you can use the `--minimal` option.
## `show-server`
You can show the configuration parameters for a given server with:
``` bash
barman show-server
```
> **TIP:** you can request a full configuration report using `all` as
> the server name.
## `status`
The `status` command shows live information and status of a PostgreSQL
server or of all servers if you use `all` as server name.
``` bash
barman show-server
```
## `switch-xlog`
This command makes the PostgreSQL server switch to another transaction
log file, allowing the current log file to be closed, received and then
archived.
``` bash
barman switch-xlog
```
If there has been no transaction activity since the last transaction
log file switch, the switch needs to be forced using the
`--force` option.
The `--archive` option requests Barman to trigger WAL archiving after
the xlog switch. By default, a 30 seconds timeout is enforced (this
can be changed with `--archive-timeout`). If no WAL file is received,
an error is returned.
barman-2.1/doc/manual/02-before_you_start.en.md 0000664 0001750 0001750 00000002053 13015546106 021752 0 ustar gcalacoci gcalacoci \newpage
# Before you start
Before you start using Barman, it is fundamental that you get familiar
with PostgreSQL and the concepts around physical backups, Point-In-Time-Recovery and replication, such as base backups, WAL archiving, etc.
Below you can find a non exhaustive list of resources that we recommend for you to read:
- _PostgreSQL documentation_:
- [SQL Dump] [sqldump] [^pgdump]
- [File System Level Backup] [physicalbackup]
- [Continuous Archiving and Point-in-Time Recovery (PITR)] [pitr]
- [Recovery Configuration] [recoveryconfig]
- [Reliability and the Write-Ahead Log] [wal]
- _Book_: [PostgreSQL 9 Administration Cookbook - 2nd edition] [adminbook]
[^pgdump]: It is important that you know the difference between logical and physical backup, therefore between `pg_dump` and a tool like Barman.
Professional training on these topics is another effective way of
learning these concepts. At any time of the year you can find many
courses available all over the world, delivered by PostgreSQL
companies such as 2ndQuadrant.
barman-2.1/doc/manual/99-references.en.md 0000664 0001750 0001750 00000005407 13015546106 020546 0 ustar gcalacoci gcalacoci
[rpo]: https://en.wikipedia.org/wiki/Recovery_point_objective
[rto]: https://en.wikipedia.org/wiki/Recovery_time_objective
[repmgr]: http://www.repmgr.org/
[sqldump]: https://www.postgresql.org/docs/current/static/backup-dump.html
[recoveryconfig]: https://www.postgresql.org/docs/current/static/recovery-config.html
[physicalbackup]: https://www.postgresql.org/docs/current/static/backup-file.html
[pitr]: https://www.postgresql.org/docs/current/static/continuous-archiving.html
[adminbook]: http://www.2ndquadrant.com/en/books/postgresql-9-administration-cookbook/
[wal]: https://www.postgresql.org/docs/current/static/wal.html
[49340627f9821e447f135455d942f7d5e96cae6d]: https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=49340627f9821e447f135455d942f7d5e96cae6d
[requirements_recovery]: https://www.postgresql.org/docs/current/static/warm-standby.html#STANDBY-PLANNING
[yumpgdg]: http://yum.postgresql.org/
[aptpgdg]: http://apt.postgresql.org/
[aptpgdgwiki]: https://wiki.postgresql.org/wiki/Apt
[epel]: http://fedoraproject.org/wiki/EPEL
[man5]: http://docs.pgbarman.org/barman.5.html
[setup_user]: https://docs.python.org/3/install/index.html#alternate-installation-the-user-scheme
[pypi]: https://pypi.python.org/pypi/barman/
[pgpass]: https://www.postgresql.org/docs/current/static/libpq-pgpass.html
[pghba]: http://www.postgresql.org/docs/current/static/client-authentication.html
[authpghba]: http://www.postgresql.org/docs/current/static/auth-pg-hba-conf.html
[streamprot]: http://www.postgresql.org/docs/current/static/protocol-replication.html
[roles]: http://www.postgresql.org/docs/current/static/role-attributes.html
[replication-slots]: https://www.postgresql.org/docs/current/static/warm-standby.html#STREAMING-REPLICATION-SLOTS
[synch]: http://www.postgresql.org/docs/current/static/warm-standby.html#SYNCHRONOUS-REPLICATION
[3]: https://sourceforge.net/projects/pgbarman/files/
[8]: http://en.wikipedia.org/wiki/Hard_link
[9]: https://github.com/2ndquadrant-it/pgespresso
[11]: http://www.pgbarman.org/
[12]: http://www.pgbarman.org/support/
[13]: http://www.2ndquadrant.com/
[14]: http://www.pgbarman.org/faq/
[15]: http://blog.2ndquadrant.com/tag/barman/
[16]: https://github.com/hamann/check-barman
[17]: https://github.com/2ndquadrant-it/puppet-barman
[18]: http://4caast.morfeo-project.org/
[19]: http://www.2ndquadrant.it/
[20]: http://www.postgresql.org/docs/current/static/functions-admin.html
[24]: http://www.postgresql.org/docs/current/static/warm-standby.html#STREAMING-REPLICATION
[25]: http://www.postgresql.org/docs/current/static/app-pgreceivexlog.html
[26]: https://goo.gl/218Ghl
[27]: https://github.com/emin100/barmanapi
[31]: http://www.postgresql.org/
barman-2.1/doc/manual/15-system_requirements.en.md 0000664 0001750 0001750 00000004520 13015546106 022533 0 ustar gcalacoci gcalacoci \newpage
# System requirements
- Linux/Unix
- Python 2.6 or 2.7
- Python modules:
- argcomplete
- argh >= 0.21.2 <= 0.26.2
- argparse (Python 2.6 only)
- psycopg2 >= 2.4.2
- python-dateutil <> 2.0
- setuptools
- PostgreSQL >= 8.3
- rsync >= 3.0.4 (optional for PostgreSQL >= 9.2)
> **IMPORTANT:**
> Users of RedHat Enterprise Linux, CentOS and Scientific Linux are
> required to install the
> [Extra Packages Enterprise Linux (EPEL) repository] [epel].
> **NOTE:**
> Python 3 support is experimental. Report any bug through
> the ticketing system on Github or the mailing list.
## Requirements for backup
The most critical requirement for a Barman server is the amount of disk space available.
You are recommended to plan the required disk space based on the size of the cluster, number of WAL files generated per day, frequency of backups, and retention policies.
Although the only file systems that we officially support are XFS and Ext4, we are aware of users that deploy Barman on different file systems including ZFS and NFS.
## Requirements for recovery
Barman allows you to recover a PostgreSQL instance either
locally (where Barman resides) or remotely (on a separate server).
Remote recovery is definitely the most common way to restore a PostgreSQL
server with Barman.
Either way, the same [requirements for PostgreSQL's Log shipping and Point-In-Time-Recovery apply] [requirements_recovery]:
- identical hardware architecture
- identical major version of PostgreSQL
In general, it is **highly recommended** to create recovery environments that are as similar as possible, if not identical, to the original server, because they are easier to maintain. For example, we suggest that you use the same operating system, the same PostgreSQL version, the same disk layouts, and so on.
Additionally, dedicated recovery environments for each PostgreSQL server, even on demand, allows you to nurture the disaster recovery culture in your team. You can be prepared for when something unexpected happens by practising
recovery operations and becoming familiar with them.
Based on our experience, designated recovery environments reduce the impact of stress in real failure situations, and therefore increase the effectiveness of recovery operations.
Finally, it is important that time is synchronised between the servers, using NTP for example.
barman-2.1/doc/manual/16-installation.en.md 0000664 0001750 0001750 00000006267 13015546106 021120 0 ustar gcalacoci gcalacoci \newpage
# Installation
> **IMPORTANT:**
> The recommended way to install Barman is by using the available
> packages for your GNU/Linux distribution.
## Installation on RedHat/CentOS using RPM packages
Barman can be installed on RHEL7, RHEL6 and RHEL5 Linux systems using
RPM packages. It is required to install the Extra Packages Enterprise
Linux (EPEL) repository beforehand.
RPM packages for Barman are available via Yum through the
[PostgreSQL Global Development Group RPM repository] [yumpgdg].
You need to follow the instructions for your distribution (for example RedHat,
CentOS, or Fedora) and architecture as detailed at
[yum.postgresql.org] [yumpgdg].
Then, as `root` simply type:
``` bash
yum install barman
```
2ndQuadrant also maintains RPM packages for Barman and distributes
them through [Sourceforge.net] [3].
## Installation on Debian/Ubuntu using packages
Barman can be installed on Debian and Ubuntu Linux systems using
packages.
It is directly available in the official repository for Debian and Ubuntu, however, these repositories might not contain the latest available version.
If you want to have the latest version of Barman, the recommended method is to install it through the [PostgreSQL Community APT repository] [aptpgdg].
Instructions can be found in the [APT section of the PostgreSQL Wiki] [aptpgdgwiki].
> **NOTE:**
> Thanks to the direct involvement of Barman developers in the
> PostgreSQL Community APT repository project, you will always have access
> to the most updated versions of Barman.
Installing Barman is as easy. As `root` user simply type:
``` bash
apt-get install barman
```
## Installation from sources
> **WARNING:**
> Manual installation of Barman from sources should only be performed
> by expert GNU/Linux users. Installing Barman this way requires
> system administration activities such as dependencies management,
> `barman` user creation, configuration of the `barman.conf` file,
> cron setup for the `barman cron` command, log management, and so on.
Create a system user called `barman` on the `backup` server.
As `barman` user, download the sources and uncompress them.
For a system-wide installation, type:
``` bash
barman@backup$ ./setup.py build
# run this command with root privileges or through sudo
barman@backup# ./setup.py install
```
For a local installation, type:
``` bash
barman@backup$ ./setup.py install --user
```
The `barman` application will be installed in your user directory ([make sure that your `PATH` environment variable is set properly] [setup_user]).
[Barman is also available on the Python Package Index (PyPI)] [pypi] and can be installed through `pip`.
## Upgrading from Barman 1.X
Version 2.0 requires that users explicitly configure
their archiving strategy. Before, the file based
archiver, controlled by `archiver`, was enabled by default.
When you upgrade your Barman installation to 2.0, make sure
you add the following line either globally or for any server
that requires it:
``` ini
archiver = on
```
Additionally, for a few releases, Barman will transparently set
`archiver = on` with any server that has not explicitly set
an archiving strategy and emit a warning.
Besides that, version 2.0 is fully compatible with older ones.
barman-2.1/doc/manual/27-windows-support.en.md 0000664 0001750 0001750 00000002467 13015546106 021623 0 ustar gcalacoci gcalacoci ## How to setup a Windows based server
You can backup a PostgreSQL server running on Windows using the
streaming connection for both WAL archiving and for backups.
> **IMPORTANT:** This feature is still experimental because it is not
> yet part of our continuous integration system.
Follow every step discussed previously for a streaming connection
setup.
> **WARNING:**: At this moment, `pg_basebackup` interoperability from
> Windows to Linux is still experimental. If you are having issues
> taking a backup from a Windows server and your PostgreSQL locale is
> not in English, a possible workaround for the issue is instructing
> your PostgreSQL to emit messages in English. You can do this by
> putting the following parameter in your `postgresql.conf` file:
>
> ``` ini
> lc_messages = 'English'
> ```
>
> This has been reported to fix the issue.
You can backup your server as usual.
Remote recovery is not supported for Windows servers, so you must
recover your cluster locally in the Barman server and then copy all
the files on a Windows server or use a folder shared between the
PostgreSQL server and the Barman server.
Additionally, make sure that the system user chosen to run PostgreSQL
has the permission needed to access the restored data. Basically, it
must have full control over the PostgreSQL data directory.
barman-2.1/doc/manual/65-troubleshooting.en.md 0000664 0001750 0001750 00000003305 13015546106 021640 0 ustar gcalacoci gcalacoci \newpage
# Troubleshooting
## Diagnose a Barman installation
You can gather important information about the status of all
the configured servers using:
``` bash
barman diagnose
```
The `diagnose` command output is a full snapshot of the barman server,
providing useful information, such as global configuration, SSH version,
Python version, `rsync` version, PostgreSQL clients version,
as well as current configuration and status of all servers.
The `diagnose` command is extremely useful for troubleshooting problems,
as it gives a global view on the status of your Barman installation.
## Requesting help
Although Barman is extensively documented, there are a lot of scenarios that
are not covered.
For any questions about Barman and disaster recovery scenarios using Barman,
you can reach the dev team using the community mailing list:
https://groups.google.com/group/pgbarman
or the IRC channel on freenode:
irc://irc.freenode.net/barman
In the event you discover a bug, you can open a ticket using Github:
https://github.com/2ndquadrant-it/barman/issues
2ndQuadrant provides professional support for Barman, including 24/7 service.
### Submitting a bug
Barman has been extensively tested and is currently being used in
several production environments. However, as any software, Barman is
not bug free.
If you discover a bug, please follow this procedure:
- execute the `barman diagnose` command
- file a bug through the Github issue tracker, by attaching the
output obtained by the diagnostics command above (`barman
diagnose`)
> **WARNING:**
> Be careful when submitting the output of the diagnose command
> as it might disclose information that are potentially dangerous
> from a security point of view.
barman-2.1/doc/manual/21-preliminary_steps.en.md 0000664 0001750 0001750 00000017126 13015546106 022160 0 ustar gcalacoci gcalacoci ## Preliminary steps
This section contains some preliminary steps that you need to
undertake before setting up your PostgreSQL server in Barman.
> **IMPORTANT:**
> Before you proceed, it is important that you have made your decision
> in terms of WAL archiving and backup strategies, as outlined in the
> _"Design and architecture"_ section. In particular, you should
> decide which WAL archiving methods to use, as well as the backup
> method.
### PostgreSQL connection
You need to make sure that the `backup` server can connect to
the PostgreSQL server on `pg` as superuser. This operation is mandatory.
We recommend creating a specific user in PostgreSQL, named `barman`,
as follows:
``` bash
postgres@pg$ createuser -s -W barman
```
> **IMPORTANT:** The above command will prompt for a password,
> which you are then advised to add to the `~barman/.pgpass` file
> on the `backup` server. For further information, please refer to
> ["The Password File" section in the PostgreSQL Documentation] [pgpass].
This connection is required by Barman in order to coordinate its
activities with the server, as well as for monitoring purposes.
You can choose your favourite client authentication method among those
offered by PostgreSQL. More information can be found in the
["Client Authentication" section of the PostgreSQL Documentation] [pghba].
Make sure you test the following command before proceeding:
``` bash
barman@backup$ psql -c 'SELECT version()' -U barman -h pg postgres
```
Write down the above information (user name, host name and database
name) and keep it for later. You will need it with in the `conninfo`
option for your server configuration, like in this example:
``` ini
[pg]
; ...
conninfo = host=pg user=barman dbname=postgres
```
> **NOTE:** Barman honours the `application_name` connection option
> for PostgreSQL servers 9.0 or higher.
### PostgreSQL WAL archiving and replication
Before you proceed, you need to properly configure PostgreSQL on `pg`
to accept streaming replication connections from the Barman
server. Please read the following sections in the PostgreSQL
documentation:
- [Role attributes] [roles]
- [The pg_hba.conf file] [authpghba]
- [Setting up standby servers using streaming replication] [streamprot]
One configuration parameter that is crucially important is the
`wal_level` parameter. This parameter must be configured to ensure
that all the useful information necessary for a backup to be coherent
are included in the transaction log file.
``` ini
wal_level = 'replica'
```
For PostgreSQL versions older than 9.6, `wal_level` must be set to
`hot_standby`.
Restart the PostgreSQL server for the configuration to be refreshed.
### PostgreSQL streaming connection
If you plan to use WAL streaming or streaming backup, you need to
setup a streaming connection. We recommend creating a specific user in
PostgreSQL, named `streaming_barman`, as follows:
``` bash
postgres@pg$ createuser -W --replication streaming_barman
```
> **IMPORTANT:** The above command will prompt for a password,
> which you are then advised to add to the `~barman/.pgpass` file
> on the `backup` server. For further information, please refer to
> ["The Password File" section in the PostgreSQL Documentation] [pgpass].
You can manually verify that the streaming connection works through
the following command:
``` bash
barman@backup$ psql -U streaming_barman -h pg \
-c "IDENTIFY_SYSTEM" \
replication=1
```
> **IMPORTANT:**
> Please make sure you are able to connect via streaming replication
> before going any further.
You also need to configure the `max_wal_senders` parameter in the
PostgreSQL configuration file:
``` ini
max_wal_senders = 2
```
This option represents the maximum number of concurrent streaming
connections that the server will be allowed to manage.
Another important parameter is `max_replication_slots`, which
represents the maximum number of replication slots [^replslot94]
that the server will be allowed to manage.
This parameter is needed if you are planning to
use the streaming connection to receive WAL files over the streaming
connection:
``` ini
max_replication_slots = 2
```
[^replslot94]: Replication slots have been introduced in PostgreSQL 9.4.
See section _"WAL Streaming / Replication slots"_ for
details.
The values proposed for `max_replication_slots` and `max_wal_senders`
must be considered as examples, and the values you will use in your
actual setup must be choosen after a careful evaluation of the
architecture. Please consult the PostgreSQL documentation for
guidelines and clarifications.
### SSH connections
SSH is a protocol and a set of tools that allows you to open a remote
shell to a remote server and copy files between the server and the local
system. You can find more documentation about SSH usage in the article
["SSH Essentials"][ssh_essentials] by Digital Ocean.
SSH key exchange is a very common practice that is used to implement
secure passwordless connections between users on different machines,
and it's needed to use `rsync` for WAL archiving and for backups.
> **NOTE:**
> This procedure is not needed if you plan to use the streaming
> connection only to archive transaction logs and backup your PostgreSQL
> server.
[ssh_essentials]: https://www.digitalocean.com/community/tutorials/ssh-essentials-working-with-ssh-servers-clients-and-keys
#### SSH configuration of postgres user
Unless you have done it before, you need to create an SSH key for the
PostgreSQL user. Log in as `postgres`, in the `pg` host and type:
``` bash
postgres@pg$ ssh-keygen -t rsa
```
As this key must be used to connect from hosts without providing a
password, no passphrase should be entered during the key pair
creation.
#### SSH configuration of barman user
As in the previous paragraph, you need to create an SSH key for the
Barman user. Log in as `barman` in the `backup` host and type:
``` bash
barman@backup$ ssh-keygen -t rsa
```
For the same reason, no passphrase should be entered.
#### From PostgreSQL to Barman
The SSH connection from the PostgreSQL server to the backup server is
needed to correctly archive WAL files using the `archive_command`
setting.
To successfully connect from the PostgreSQL server to the backup
server, the PostgreSQL public key has to be configured into the
authorized keys of the backup server for the `barman` user.
The public key to be authorized is stored inside the `postgres` user
home directory in a file named `.ssh/id_rsa.pub`, and its content
should be included in a file named `.ssh/authorized_keys` inside the
home directory of the `barman` user in the backup server. If the
`authorized_keys` file doesn't exist, create it using `600` as
permissions.
The following command should succeed without any output if the SSH key
pair exchange has been completed successfully:
``` bash
postgres@pg$ ssh barman@backup -C true
```
The value of the `archive_command` configuration parameter will be
discussed in the _"WAL archiving via archive_command section"_.
#### From Barman to PostgreSQL
The SSH connection between the backup server and the PostgreSQL server
is used for the traditional backup over rsync. Just as with the
connection from the PostgreSQL server to the backup server, we should
authorize the public key of the backup server in the PostgreSQL server
for the `postgres` user.
The content of the file `.ssh/id_rsa.pub` in the `barman` server should
be put in the file named `.ssh/authorized_keys` in the PostgreSQL
server. The permissions of that file should be `600`.
The following command should succeed without any output if the key
pair exchange has been completed successfully.
``` bash
barman@backup$ ssh postgres@pg -C true
```
barman-2.1/doc/manual/01-intro.en.md 0000664 0001750 0001750 00000007661 13015546106 017543 0 ustar gcalacoci gcalacoci \newpage
# Introduction
In a perfect world, there would be no need for a backup. However, it is
important, especially in business environments, to be prepared for
when the _"unexpected"_ happens. In a database scenario, the
unexpected could take any of the following forms:
- data corruption
- system failure (including hardware failure)
- human error
- natural disaster
In such cases, any ICT manager or DBA should be able to fix the
incident and recover the database in the shortest time possible. We
normally refer to this discipline as **disaster recovery**, and more
broadly *business continuity*.
Within business continuity, it is important to familiarise with two fundamental metrics, as defined by Wikipedia:
- [**Recovery Point Objective (RPO)**] [rpo]: _"maximum targeted period in which data might be lost from an IT service due to a major incident"_
- [**Recovery Time Objective (RTO)**] [rto]: _"the targeted duration of time and a service level within which a business process must be restored after a disaster (or disruption) in order to avoid unacceptable consequences associated with a break in business continuity"_
In a few words, RPO represents the maximum amount of data you can afford to lose, while RTO represents the maximum down-time you can afford for your service.
Understandably, we all want **RPO=0** (*"zero data loss"*) and **RTO=0** (*zero down-time*, utopia) - even if it is our grandmothers's recipe website.
In reality, a careful cost analysis phase allows you to determine your business continuity requirements.
Fortunately, with an open source stack composed of **Barman** and **PostgreSQL**, you can achieve RPO=0 thanks to synchronous streaming replication. RTO is more the focus of a *High Availability* solution, like [**repmgr**] [repmgr]. Therefore, by integrating Barman and repmgr, you can dramatically reduce RTO to nearly zero.
Based on our experience at 2ndQuadrant, we can confirm that PostgreSQL open source clusters with Barman and repmgr can easily achieve more than 99.99% uptime over a year, if properly configured and monitored.
In any case, it is important for us to emphasise more on cultural aspects related to disaster recovery, rather than the actual tools. Tools without human beings are useless.
Our mission with Barman is to promote a culture of disaster recovery that:
- focuses on backup procedures
- focuses even more on recovery procedures
- relies on education and training on strong theoretical and practical concepts of PostgreSQL's crash recovery, backup, Point-In-Time-Recovery, and replication for your team members
- promotes testing your backups (only a backup that is tested can be considered to be valid), either manually or automatically (be creative with Barman's hook scripts!)
- fosters regular practice of recovery procedures, by all members of your devops team (yes, developers too, not just system administrators and DBAs)
- solicites to regularly scheduled drills and disaster recovery simulations with the team every 3-6 months
- relies on continuous monitoring of PostgreSQL and Barman, and that is able to promptly identify any anomalies
Moreover, do everything you can to prepare yourself and your team for when the disaster happens (yes, *when*), because when it happens:
- It is going to be a Friday evening, most likely right when you are about to leave the office.
- It is going to be when you are on holiday (right in the middle of your cruise around the world) and somebody else has to deal with it.
- It is certainly going to be stressful.
- You will regret not being sure that the last available backup is valid.
- Unless you know how long it approximately takes to recover, every second will seems like forever.
Be prepared, don't be scared.
In 2011, with these goals in mind, 2ndQuadrant started the development of
Barman, now one of the most used backup tools for PostgreSQL. Barman is an acronym for "Backup and Recovery Manager".
Currently, Barman works only on Linux and Unix operating systems.
barman-2.1/doc/manual/17-configuration.en.md 0000664 0001750 0001750 00000006355 13015546106 021265 0 ustar gcalacoci gcalacoci \newpage
# Configuration
There are two types of configuration files in Barman:
- **global/general configuration**
- **server configuration**
The main configuration file (set to `/etc/barman.conf` by default) contains general options such as main directory, system user, log file, and so on.
Server configuration files, one for each server to be backed up by Barman, are located in the `/etc/barman.d` directory and must have a `.conf` suffix.
> **IMPORTANT**: For historical reasons, you can still have one single
> configuration file containing both global and server options. However,
> for maintenance reasons, this approach is deprecated.
Configuration files in Barman follow the _INI_ format.
## Options scope
Every configuration option has a _scope_:
- global
- server
- global/server: server options that can be generally set at global level
Global options are allowed in the _general section_, which is identified in the INI file by the `[barman]` label:
``` ini
[barman]
; ... global and global/server options go here
```
Server options can only be specified in a _server section_, which is identified by a line in the configuration file, in square brackets (`[` and `]`). The server section represents the ID of that server in Barman. The following example specifies a section for the server named `pg`:
``` ini
[pg]
; Configuration options for the
; server named 'pg' go here
```
There are two reserved words that cannot be used as server names in Barman:
- `barman`: identifier of the global section
- `all`: a handy shortcut that allows you to execute some commands on every server managed by Barman in sequence
Barman implements the **convention over configuration** design paradigm, which attempts to reduce the number of options that you are required to configure without losing flexibility. Therefore, some server options can be defined at global level and overridden at server level, allowing users to specify a generic behavior and refine it for one or more servers. These options have a global/server scope.
For a list of all the available configurations
and their scope, please refer to [section 5 of the 'man' page] [man5].
``` bash
man 5 barman
```
## Examples of configuration
The following is a basic example of main configuration file:
``` ini
[barman]
barman_user = barman
configuration_files_directory = /etc/barman.d
barman_home = /var/lib/barman
log_file = /var/log/barman/barman.log
log_level = INFO
compression = gzip
```
The example below, on the other hand, is a server configuration file that uses streaming backup:
``` ini
[streaming-pg]
description = "Example of PostgreSQL Database (Streaming-Only)"
conninfo = host=pg user=barman dbname=postgres
streaming_conninfo = host=pg user=streaming_barman
backup_method = postgres
streaming_archiver = on
slot_name = barman
```
The following code shows a basic example of traditional backup using `rsync`/SSH:
``` ini
[ssh-pg]
description = "Example of PostgreSQL Database (via Ssh)"
ssh_command = ssh postgres@pg
conninfo = host=pg user=barman dbname=postgres
backup_method = rsync
reuse_backup = link
archiver = on
```
For more detailed information, please refer to the distributed
`barman.conf` file, as well as the `ssh-server.conf-template` and `streaming-server.conf-template` template files.
barman-2.1/doc/manual/50-feature-details.en.md 0000664 0001750 0001750 00000052336 13015546106 021471 0 ustar gcalacoci gcalacoci \newpage
# Features in detail
In this section we present several Barman features and discuss their
applicability and the configuration required to use them.
This list is not exhaustive, as many scenarios can be created working
on the Barman configuration. Nevertheless, it is useful to discuss
common patterns.
## Backup features
### Incremental backup
Barman implements **file-level incremental backup**. Incremental
backup is a type of full periodic backup which only saves data changes
from the latest full backup available in the catalog for a specific
PostgreSQL server. It must not be confused with differential backup,
which is implemented by _WAL continuous archiving_.
> **NOTE:** Block level incremental backup will be available in
> future versions.
> **IMPORTANT:** The `reuse_backup` option can't be used with the
> `postgres` backup method at this time.
The main goals of incremental backups in Barman are:
- Reduce the time taken for the full backup process
- Reduce the disk space occupied by several periodic backups (**data
deduplication**)
This feature heavily relies on `rsync` and [hard links] [8], which
must therefore be supported by both the underlying operating system
and the file system where the backup data resides.
The main concept is that a subsequent base backup will share those
files that have not changed since the previous backup, leading to
relevant savings in disk usage. This is particularly true of VLDB
contexts and of those databases containing a high percentage of
_read-only historical tables_.
Barman implements incremental backup through a global/server option
called `reuse_backup`, that transparently manages the `barman backup`
command. It accepts three values:
- `off`: standard full backup (default)
- `link`: incremental backup, by reusing the last backup for a server
and creating a hard link of the unchanged files (for backup space
and time reduction)
- `copy`: incremental backup, by reusing the last backup for a server
and creating a copy of the unchanged files (just for backup time
reduction)
The most common scenario is to set `reuse_backup` to `link`, as
follows:
``` ini
reuse_backup = link
```
Setting this at global level will automatically enable incremental
backup for all your servers.
As a final note, users can override the setting of the `reuse_backup`
option through the `--reuse-backup` runtime option for the `barman
backup` command. Similarly, the runtime option accepts three values:
`off`, `link` and `copy`. For example, you can run a one-off
incremental backup as follows:
``` bash
barman backup --reuse-backup=link
```
### Limiting bandwidth usage
It is possible to limit the usage of I/O bandwidth through the
`bandwidth_limit` option (global/per server), by specifying the
maximum number of kilobytes per second. By default it is set to 0,
meaning no limit.
> **IMPORTANT:** the `bandwidth_limit` and the
> `tablespace_bandwidth_limit` options are not supported with the
> `postgres` backup method
In case you have several tablespaces and you prefer to limit the I/O
workload of your backup procedures on one or more tablespaces, you can
use the `tablespace_bandwidth_limit` option (global/per server):
``` ini
tablespace_bandwidth_limit = tbname:bwlimit[, tbname:bwlimit, ...]
```
The option accepts a comma separated list of pairs made up of the
tablespace name and the bandwidth limit (in kilobytes per second).
When backing up a server, Barman will try and locate any existing
tablespace in the above option. If found, the specified bandwidth
limit will be enforced. If not, the default bandwidth limit for that
server will be applied.
### Network Compression
It is possible to reduce the size of transferred data using
compression. It can be enabled using the `network_compression` option
(global/per server):
> **IMPORTANT:** the `network_compression` option is not available
> with the `postgres` backup method.
``` ini
network_compression = true|false
```
Setting this option to `true` will enable data compression during
network transfers (for both backup and recovery). By default it is set
to `false`.
### Concurrent Backup and backup from a standby
Normally, during backup operations, Barman uses PostgreSQL native
functions `pg_start_backup` and `pg_stop_backup` for _exclusive
backup_. These operations are not allowed on a read-only standby
server.
Barman is also capable of performing backups of PostgreSQL from 9.2 or
greater database servers in a **concurrent way**, primarily through
the `backup_options` configuration
parameter.[^ABOUT_CONCURRENT_BACKUP]
[^ABOUT_CONCURRENT_BACKUP]:
Concurrent backup is a technology that has been available in
PostgreSQL since version 9.2, through the _streaming replication
protocol_ (for example, using a tool like `pg_basebackup`).
This introduces a new architecture scenario with Barman: **backup from
a standby server**, using `rsync`.
> **IMPORTANT:** **Concurrent backup** requires users of PostgreSQL
> 9.2, 9.3, 9.4, and 9.5 to install the `pgespresso` open source
> extension on every PostgreSQL server of the cluster. For more
> detailed information and the source code, please visit the
> [pgespresso extension website] [9]. Barman supports the new API
> introduced in PostgreSQL 9.6. This removes the requirement of the
> `pgespresso` extension to perform concurrent backups from this
> version of PostgreSQL.
By default, `backup_options` is transparently set to
`exclusive_backup` for back compatibility reasons.
Users of PostgreSQL 9.6 should set `backup_options` to `concurrent_backup`.
When `backup_options` is set to `concurrent_backup`, Barman activates
the _concurrent backup mode_ for a server and follows these two simple
rules:
- `ssh_command` must point to the destination Postgres server
- `conninfo` must point to a database on the destination Postgres
database. Using PostgreSQL 9.2, 9.3, 9.4, and 9.5, `pgespresso`
must be correctly installed through `CREATE EXTENSION`. Using 9.6 or
greater, concurrent backups are executed through the Postgres native
API.
The destination Postgres server can be either the master or a
streaming replicated standby server.
> **NOTE:**
> When backing up from a standby server, continuous archiving of WAL
> files must be configured on the master to ship files to the Barman
> server (as outlined in the _"WAL archiving via archive_command"_ section
> above)[^CONCURRENT_ARCHIVING].
[^CONCURRENT_ARCHIVING]:
In case of a concurrent backup, currently Barman has no way
to determine that the closing WAL file of a full backup has
actually been shipped - opposite of an exclusive backup
where PostgreSQL itself makes sure that the WAL file is correctly
archived. Be aware that the full backup cannot be considered
consistent until that WAL file has been received and archived by
Barman.
## Archiving features
### WAL compression
The `barman cron` command will compress WAL files if the `compression`
option is set in the configuration file. This option allows five
values:
- `bzip2`: for Bzip2 compression (requires the `bzip2` utility)
- `gzip`: for Gzip compression (requires the `gzip` utility)
- `pybzip2`: for Bzip2 compression (uses Python's internal compression module)
- `pygzip`: for Gzip compression (uses Python's internal compression module)
- `pigz`: for Pigz compression (requires the `pigz` utility)
- `custom`: for custom compression, which requires you to set the
following options as well:
- `custom_compression_filter`: a compression filter
- `custom_decompression_filter`: a decompression filter
> *NOTE:* All methods but `pybzip2` and `pygzip` require `barman
> archive-wal` to fork a new process.
### Synchronous WAL streaming
> **IMPORTANT:** This feature is available only from PostgreSQL 9.5
> and above.
Barman can also reduce the Recovery Point Objective to zero, by
collecting the transaction WAL files like a synchronous standby server
would.
To configure such a scenario, the Barman server must be configured to
archive WALs via the [streaming connection](#streaming_connection),
and the `receive-wal` process should figure as a synchronous standby
of the PostgreSQL server.
First of all, you need to retrieve the application name of the Barman
`receive-wal` process with the `show-server` command:
``` bash
barman@backup$ barman show-server pg|grep streaming_archiver_name
streaming_archiver_name: barman_receive_wal
```
Then the application name should be added to the `postgresql.conf`
file as a synchronous standby:
``` ini
synchronous_standby_names = 'barman_receive_wal'
```
> **IMPORTANT:** this is only an example of configuration, to show you that
> Barman is eligible to be a synchronous standby node.
> We are not suggesting to use ONLY Barman.
> You can read _["Synchronous Replication"][synch]_ from the PostgreSQL
> documentation for further information on this topic.
The PostgreSQL server needs to be restarted for the configuration to
be reloaded.
If the server has been configured correctly, the `replication-status`
command should show the `receive_wal` process as a synchronous
streaming client:
``` bash
[root@backup ~]# barman replication-status pg
Status of streaming clients for server 'pg':
Current xlog location on master: 0/9000098
Number of streaming clients: 1
1. #1 Sync WAL streamer
Application name: barman_receive_wal
Sync stage : 3/3 Remote write
Communication : TCP/IP
IP Address : 139.59.135.32 / Port: 58262 / Host: -
User name : streaming_barman
Current state : streaming (sync)
Replication slot: barman
WAL sender PID : 2501
Started at : 2016-09-16 10:33:01.725883+00:00
Sent location : 0/9000098 (diff: 0 B)
Write location : 0/9000098 (diff: 0 B)
Flush location : 0/9000098 (diff: 0 B)
```
## Catalog management features
### Minimum redundancy safety
You can define the minimum number of periodic backups for a PostgreSQL
server, using the global/per server configuration option called
`minimum_redundancy`, by default set to 0.
By setting this value to any number greater than 0, Barman makes sure
that at any time you will have at least that number of backups in a
server catalog.
This will protect you from accidental `barman delete` operations.
> **IMPORTANT:**
> Make sure that your retention policy settings do not collide with
> minimum redundancy requirements. Regularly check Barman's log for
> messages on this topic.
### Retention policies
Barman supports **retention policies** for backups.
A backup retention policy is a user-defined policy that determines how
long backups and related archive logs (Write Ahead Log segments) need
to be retained for recovery procedures.
Based on the user's request, Barman retains the periodic backups
required to satisfy the current retention policy and any archived WAL
files required for the complete recovery of those backups.
Barman users can define a retention policy in terms of **backup
redundancy** (how many periodic backups) or a **recovery window** (how
long).
Retention policy based on redundancy
: In a redundancy based retention policy, the user determines how
many periodic backups to keep. A redundancy-based retention policy
is contrasted with retention policies that use a recovery window.
Retention policy based on recovery window
: A recovery window is one type of Barman backup retention policy,
in which the DBA specifies a period of time and Barman ensures
retention of backups and/or archived WAL files required for
point-in-time recovery to any time during the recovery window. The
interval always ends with the current time and extends back in
time for the number of days specified by the user. For example, if
the retention policy is set for a recovery window of seven days,
and the current time is 9:30 AM on Friday, Barman retains the
backups required to allow point-in-time recovery back to 9:30 AM
on the previous Friday.
#### Scope
Retention policies can be defined for:
- **PostgreSQL periodic base backups**: through the `retention_policy`
configuration option
- **Archive logs**, for Point-In-Time-Recovery: through the
`wal_retention_policy` configuration option
> **IMPORTANT:**
> In a temporal dimension, archive logs must be included in the time
> window of periodic backups.
There are two typical use cases here: full or partial point-in-time
recovery.
Full point in time recovery scenario:
: Base backups and archive logs share the same retention policy,
allowing you to recover at any point in time from the first
available backup.
Partial point in time recovery scenario:
: Base backup retention policy is wider than that of archive logs,
for example allowing users to keep full, weekly backups of the
last 6 months, but archive logs for the last 4 weeks (granting to
recover at any point in time starting from the last 4 periodic
weekly backups).
> **IMPORTANT:**
> Currently, Barman implements only the **full point in time
> recovery** scenario, by constraining the `wal_retention_policy`
> option to `main`.
#### How they work
Retention policies in Barman can be:
- **automated**: enforced by `barman cron`
- **manual**: Barman simply reports obsolete backups and allows you
to delete them
> **IMPORTANT:**
> Currently Barman does not implement manual enforcement. This feature
> will be available in future versions.
#### Configuration and syntax
Retention policies can be defined through the following configuration
options:
- `retention_policy`: for base backup retention
- `wal_retention_policy`: for archive logs retention
- `retention_policy_mode`: can only be set to `auto` (retention
policies are automatically enforced by the `barman cron` command)
These configuration options can be defined both at a global level and
a server level, allowing users maximum flexibility on a multi-server
environment.
##### Syntax for `retention_policy`
The general syntax for a base backup retention policy through
`retention_policy` is the following:
``` ini
retention_policy = {REDUNDANCY value | RECOVERY WINDOW OF value {DAYS | WEEKS | MONTHS}}
```
Where:
- syntax is case insensitive
- `value` is an integer and is > 0
- in case of **redundancy retention policy**:
- `value` must be greater than or equal to the server minimum
redundancy level (if that value is not assigned,
a warning is generated)
- the first valid backup is the value-th backup in a reverse
ordered time series
- in case of **recovery window policy**:
- the point of recoverability is: current time - window
- the first valid backup is the first available backup before
the point of recoverability; its value in a reverse ordered
time series must be greater than or equal to the server
minimum redundancy level (if it is not assigned to that value
and a warning is generated)
By default, `retention_policy` is empty (no retention enforced).
##### Syntax for `wal_retention_policy`
Currently, the only allowed value for `wal_retention_policy` is the
special value `main`, that maps the retention policy of archive logs
to that of base backups.
## Hook scripts
Barman allows a database administrator to run hook scripts on these
two events:
- before and after a backup
- before and after a WAL file is archived
There are two types of hook scripts that Barman can manage:
- standard hook scripts
- retry hook scripts
The only difference between these two types of hook scripts is that
Barman executes a standard hook script only once, without checking its
return code, whereas a retry hook script may be executed more than
once, depending on its return code.
Specifically, when executing a retry hook script, Barman checks the
return code and retries indefinitely until the script returns either
`SUCCESS` (with standard return code `0`), or `ABORT_CONTINUE` (return
code `62`), or `ABORT_STOP` (return code `63`). Barman treats any
other return code as a transient failure to be retried. Users are
given more power: a hook script can control its workflow by specifying
whether a failure is transient. Also, in case of a 'pre' hook script,
by returning `ABORT_STOP`, users can request Barman to interrupt the
main operation with a failure.
Hook scripts are executed in the following order:
1. The standard 'pre' hook script (if present)
2. The retry 'pre' hook script (if present)
3. The actual event (i.e. backup operation, or WAL archiving), if
retry 'pre' hook script was not aborted with `ABORT_STOP`
4. The retry 'post' hook script (if present)
5. The standard 'post' hook script (if present)
The output generated by any hook script is written in the log file of
Barman.
> **NOTE:**
> Currently, `ABORT_STOP` is ignored by retry 'post' hook scripts. In
> these cases, apart from lodging an additional warning, `ABORT_STOP`
> will behave like `ABORT_CONTINUE`.
### Backup scripts
These scripts can be configured with the following global
configuration options (which can be overridden on a per server basis):
- `pre_backup_script`: _hook script_ executed _before_ a base backup,
only once, with no check on the exit code
- `pre_backup_retry_script`: _retry hook script_ executed _before_ a
base backup, repeatedly until success or abort
- `post_backup_retry_script`: _retry hook script_ executed _after_ a
base backup, repeatedly until success or abort
- `post_backup_script`: _hook script_ executed _after_ a base backup,
only once, with no check on the exit code
The script definition is passed to a shell and can return any exit
code. Only in case of a _retry_ script, Barman checks the return code
(see the [hook script section](#hook_scripts)).
The shell environment will contain the following variables:
- `BARMAN_BACKUP_DIR`: backup destination directory
- `BARMAN_BACKUP_ID`: ID of the backup
- `BARMAN_CONFIGURATION`: configuration file used by Barman
- `BARMAN_ERROR`: error message, if any (only for the `post` phase)
- `BARMAN_PHASE`: phase of the script, either `pre` or `post`
- `BARMAN_PREVIOUS_ID`: ID of the previous backup (if present)
- `BARMAN_RETRY`: `1` if it is a retry script, `0` if not
- `BARMAN_SERVER`: name of the server
- `BARMAN_STATUS`: status of the backup
- `BARMAN_VERSION`: version of Barman
### WAL archive scripts
Similar to backup scripts, archive scripts can be configured with
global configuration options (which can be overridden on a per server
basis):
- `pre_archive_script`: _hook script_ executed _before_ a WAL file is
archived by maintenance (usually `barman cron`), only once, with no
check on the exit code
- `pre_archive_retry_script`: _retry hook script_ executed _before_ a
WAL file is archived by maintenance (usually `barman cron`),
repeatedly until it is successful or aborted
- `post_archive_retry_script`: _retry hook script_ executed _after_ a
WAL file is archived by maintenance, repeatedly until it is
successful or aborted
- `post_archive_script`: _hook script_ executed _after_ a WAL file is
archived by maintenance, only once, with no check on the exit code
The script is executed through a shell and can return any exit code.
Only in case of a _retry_ script, Barman checks the return code (see
the upper section).
Archive scripts share with backup scripts some environmental
variables:
- `BARMAN_CONFIGURATION`: configuration file used by Barman
- `BARMAN_ERROR`: error message, if any (only for the `post` phase)
- `BARMAN_PHASE`: phase of the script, either `pre` or `post`
- `BARMAN_SERVER`: name of the server
Following variables are specific to archive scripts:
- `BARMAN_SEGMENT`: name of the WAL file
- `BARMAN_FILE`: full path of the WAL file
- `BARMAN_SIZE`: size of the WAL file
- `BARMAN_TIMESTAMP`: WAL file timestamp
- `BARMAN_COMPRESSION`: type of compression used for the WAL file
## Customization
### Lock file directory
Barman allows you to specify a directory for lock files through the
`barman_lock_directory` global option.
Lock files are used to coordinate concurrent work at global and server
level (for example, cron operations, backup operations, access to the
WAL archive, and so on.).
By default (for backward compatibility reasons),
`barman_lock_directory` is set to `barman_home`.
> **TIP:**
> Users are encouraged to use a directory in a volatile partition,
> such as the one dedicated to run-time variable data (e.g.
> `/var/run/barman`).
### Binary paths
As of version 1.6.0, Barman allows users to specify one or more directories
where Barman looks for executable files, using the global/server
option `path_prefix`.
If a `path_prefix` is provided, it must contain a list of one or more
directories separated by colon. Barman will search inside these directories
first, then in those specified by the `PATH` environment variable.
By default the `path_prefix` option is empty.
## Integration with cluster management systems
Barman has been designed for integration with standby servers (with
streaming replication or traditional file based log shipping) and high
availability tools like [repmgr] [repmgr].
From an architectural point of view, PostgreSQL must be configured to
archive WAL files directly to the Barman server.
Barman, thanks to the `get-wal` framework, can also be used as a WAL hub.
For this purpose, you can use the `barman-wal-restore` script, part
of the `barman-cli` package, with all your standby servers.
The `replication-status` command allows
you to get information about any streaming client attached to the
managed server, in particular hot standby servers and WAL streamers.
barman-2.1/doc/manual/26-rsync_backup.en.md 0000664 0001750 0001750 00000001720 13015546106 021070 0 ustar gcalacoci gcalacoci ## Backup with `rsync`/SSH
The backup over `rsync` was the only available method before 2.0, and
is currently the only backup method that supports the incremental
backup feature. Please consult the _"Features in detail"_ section for
more information.
To take a backup using `rsync` you need to put these parameters inside
the Barman server configuration file:
``` ini
backup_method = rsync
ssh_command = ssh postgres@pg
```
The `backup_method` option activates the `rsync` backup method, and
the `ssh_command` option is needed to correctly create an SSH
connection from the Barman server to the PostgreSQL server.
> **IMPORTANT:** Keep in mind that if the WAL archiving is not
> currently configured, you will not be able to start a backup.
To check if the server configuration is valid you can use the `barman
check` command:
``` bash
barman@backup$ barman check pg
```
To take a backup use the `barman backup` command:
``` bash
barman@backup$ barman backup pg
```
barman-2.1/doc/manual/23-wal_streaming.en.md 0000664 0001750 0001750 00000007230 13015546106 021240 0 ustar gcalacoci gcalacoci ## WAL streaming
Barman can reduce the Recovery Point Objective (RPO) by allowing users
to add continuous WAL streaming from a PostgreSQL server, on top of
the standard `archive_command` strategy
Barman relies on [`pg_receivexlog`] [25], a utility that has been
available from PostgreSQL 9.2 which exploits the native streaming
replication protocol and continuously receives transaction logs from a
PostgreSQL server (master or standby).
> **IMPORTANT:**
> Barman requires that `pg_receivexlog` is installed on the same
> server. For PostgreSQL 9.2 servers, you need `pg_receivexlog` of
> version 9.2 installed alongside Barman. For PostgreSQL 9.3 and
> above, it is recommended to install the latest available version of
> `pg_receivexlog`, as it is back compatible. Otherwise, users can
> install multiple versions of `pg_receivexlog` on the Barman server
> and properly point to the specific version for a server, using the
> `path_prefix` option in the configuration file.
In order to enable streaming of transaction logs, you need to:
1. setup a streaming connection as previously described
2. set the `streaming_archiver` option to `on`
The `cron` command, if the aforementioned requirements are met,
transparently manages log streaming through the execution of the
`receive-wal` command. This is the recommended scenario.
However, users can manually execute the `receive-wal` command:
``` bash
barman receive-wal
```
> **NOTE:**
> The `receive-wal` command is a foreground process.
Transaction logs are streamed directly in the directory specified by the
`streaming_wals_directory` configuration option and are then archived
by the `archive-wal` command.
Unless otherwise specified in the `streaming_archiver_name` parameter,
and only for PostgreSQL 9.3 or above, Barman will set `application_name`
of the WAL streamer process to `barman_receive_wal`, allowing you to
monitor its status in the `pg_stat_replication` system view of the
PostgreSQL server.
### Replication slots
> **IMPORTANT:** replication slots are available since PostgreSQL 9.4
Replication slots are an automated way to ensure that the PostgreSQL
server will not remove WAL files until they were received by all
archivers. Barman uses this mechanism to receive the transaction logs
from PostgreSQL.
You can find more information about replication slots in the
[PostgreSQL manual][replication-slots].
You can even base your backup architecture on streaming connection
only. This scenario is useful to configure Docker-based PostgreSQL
servers and even to work with PostgreSQL servers running on Windows.
> **IMPORTANT:**
> In this moment, the Windows support is still experimental, as it is
> not yet part of our continuous integration system.
### How to configure the WAL streaming
First, the PostgreSQL server must be configured to stream the
transaction log files to the Barman server.
To configure the streaming connection from Barman to the PostgreSQL
server you need to enable the `streaming_archiver`, as already said,
including this line in the server configuration file:
``` ini
streaming_archiver = on
```
If you plan to use replication slots (recommended),
another essential option for the setup of the streaming-based
transaction log archiving is the `slot_name` option:
``` ini
slot_name = barman
```
This option defines the name of the replication slot that will be
used by Barman. It is mandatory if you want to use replication slots.
When you configure the replication slot name, you can create a
replication slot for Barman with this command:
``` bash
barman@backup$ barman receive-wal --create-slot pg
Creating physical replication slot 'barman' on server 'pg'
Replication slot 'barman' created
```
barman-2.1/doc/manual/66-about.en.md 0000664 0001750 0001750 00000007425 13015546106 017533 0 ustar gcalacoci gcalacoci \newpage
# The Barman project
## Support and sponsor opportunities
Barman is free software, written and maintained by 2ndQuadrant. If you
require support on using Barman, or if you need new features, please
get in touch with 2ndQuadrant. You can sponsor the development of new
features of Barman and PostgreSQL which will be made publicly
available as open source.
For further information, please visit:
- [Barman website] [11]
- [Support section] [12]
- [2ndQuadrant website] [13]
- [Barman FAQs] [14]
- [2ndQuadrant blog: Barman] [15]
## Contributing to Barman
2ndQuadrant has a team of software engineers, architects, database
administrators, system administrators, QA engineers, developers and
managers that dedicate their time and expertise to improve Barman's code.
We adopt lean and agile methodologies for software development, and
we believe in the _devops_ culture that allowed us to implement rigorous
testing procedures through cross-functional collaboration.
Every Barman commit is the contribution of multiple individuals, at different
stages of the production pipeline.
Even though this is our preferred way of developing Barman, we gladly
accept patches from external developers, as long as:
- user documentation (tutorial and man pages) is provided.
- source code is properly documented and contains relevant comments.
- code supplied is covered by unit tests.
- no unrelated feature is compromised or broken.
- source code is rebased on the current master branch.
- commits and pull requests are limited to a single feature (multi-feature
patches are hard to test and review).
- changes to the user interface are discussed beforehand with 2ndQuadrant.
We also require that any contributions provide a copyright assignment
and a disclaimer of any work-for-hire ownership claims from the employer
of the developer.
You can use Github's pull requests system for this purpose.
## Authors
In alphabetical order:
- Gabriele Bartolini, (project leader)
- Jonathan Battiato, (QA/testing)
- Stefano Bianucci, (developer, intern from University of Florence)
- Giuseppe Broccolo, (QA/testing)
- Giulio Calacoci, (developer)
- Francesco Canovai, (QA/testing)
- Leonardo Cecchi, (developer)
- Gianni Ciolli, (QA/testing)
- Britt Cole, (documentation)
- Marco Nenciarini, (lead developer)
- Rubens Souza, (QA/testing)
Past contributors:
- Carlo Ascani
## Links
- [check-barman] [16]: a Nagios plugin for Barman, written by Holger
Hamann (MIT license)
- [puppet-barman] [17]: Barman module for Puppet (GPL)
- [Tutorial on "How To Back Up, Restore, and Migrate PostgreSQL Databases with Barman on CentOS 7"] [26], by Sadequl Hussain (available on DigitalOcean Community)
- [BarmanAPI] [27]: RESTFul API for Barman, written by Mehmet Emin KarakaĹź (GPL)
## License and Contributions
Barman is the property of 2ndQuadrant Italia and its code is
distributed under GNU General Public License 3.
Copyright (C) 2011-2016 [2ndQuadrant.it S.r.l.] [19].
Barman has been partially funded through [4CaaSt] [18], a research
project funded by the European Commission's Seventh Framework
programme.
Contributions to Barman are welcome, and will be listed in the
`AUTHORS` file. 2ndQuadrant Italia requires that any contributions
provide a copyright assignment and a disclaimer of any work-for-hire
ownership claims from the employer of the developer. This lets us make
sure that all of the Barman distribution remains free code. Please
contact info@2ndQuadrant.it for a copy of the relevant Copyright
Assignment Form.
barman-2.1/doc/manual/22-config_file.en.md 0000664 0001750 0001750 00000001537 13015546106 020653 0 ustar gcalacoci gcalacoci ## The server configuration file
Create a new file, called `pg.conf`, in `/etc/barman.d` directory, with the following content:
``` ini
[pg]
description = "Our main PostgreSQL server"
conninfo = host=pg user=barman dbname=postgres
backup_method = postgres
# backup_method = rsync
```
The `conninfo` option is set accordingly to the section _"Preliminary
steps: PostgreSQL connection"_.
The meaning of the `backup_method` option will be covered in the
backup section of this guide.
If you plan to use the streaming connection for WAL archiving or to
create a backup of your server, you also need a `streaming_conninfo`
parameter in your server configuration file:
``` ini
streaming_conninfo = host=pg user=streaming_barman dbname=postgres
```
This value must be choosen accordingly as described in the section
_"Preliminary steps: PostgreSQL connection"_.
barman-2.1/doc/manual/20-server_setup.en.md 0000664 0001750 0001750 00000000673 13015546106 021133 0 ustar gcalacoci gcalacoci \newpage
# Setup of a new server in Barman
As mentioned in the _"Design and architecture"_ section, we will use the
following conventions:
- `pg` as server ID and host name where PostgreSQL is installed
- `backup` as host name where Barman is located
- `barman` as the user running Barman on the `backup` server (identified by
the parameter `barman_user` in the configuration)
- `postgres` as the user running PostgreSQL on the `pg` server
barman-2.1/doc/manual/70-feature-matrix.en.md 0000664 0001750 0001750 00000003617 13015546106 021350 0 ustar gcalacoci gcalacoci \newpage
\appendix
# Feature matrix
Below you will find a matrix of PostgreSQL versions and Barman features for backup and archiving:
| **Version** | **Backup with rsync/SSH** | **Backup with pg_basebackup** | **Standard WAL archiving** | **WAL Streaming** | **RPO=0** |
|:---------:|:---------------------:|:-------------------------:|:----------------------:|:----------------------:|:-------:|
| **9.6** | Yes | Yes | Yes | Yes | Yes |
| **9.5** | Yes | Yes | Yes | Yes | Yes ~(d)~ |
| **9.4** | Yes | Yes | Yes | Yes | Yes ~(d)~ |
| **9.3** | Yes | Yes ~(c)~ | Yes | Yes ~(b)~ | No |
| **9.2** | Yes | Yes ~(a)~~(c)~ | Yes | Yes ~(a)~~(b)~ | No |
| _9.1_ | Yes | No | Yes | No | No |
| _9.0_ | Yes | No | Yes | No | No |
| _8.4_ | Yes | No | Yes | No | No |
| _8.3_ | Yes | No | Yes | No | No |
**NOTE:**
a) `pg_basebackup` and `pg_receivexlog` 9.2 required
b) WAL streaming-only not supported (standard archiving required)
c) Backup of tablespaces not supported
d) When using `pg_receivexlog` 9.5, minor version 9.5.5 or higher required [^commitsync]
[^commitsync]: The commit ["Fix pg_receivexlog --synchronous"] [49340627f9821e447f135455d942f7d5e96cae6d] is required (included in version 9.5.5)
It is required by Barman that `pg_basebackup` and `pg_receivexlog` of the same version of the PostgreSQL server (or higher) are installed on the same server where Barman resides. The only exception is that PostgreSQL 9.2 users are required to install version 9.2 of `pg_basebackup` and `pg_receivexlog` alongside with Barman.
>> **TIP:** We recommend that the last major, stable version of the PostgreSQL clients (e.g. 9.6) is installed on the Barman server if you plan to use backup and WAL archiving over streaming replication through `pg_basebackup` and `pg_receivexlog`, for PostgreSQL 9.3 or higher servers.
>> **TIP:** For "RPO=0" architectures, it is recommended to have at least one synchronous standby server.
barman-2.1/doc/manual/41-global-commands.en.md 0000664 0001750 0001750 00000003511 13015546106 021441 0 ustar gcalacoci gcalacoci \newpage
# General commands
Barman has many commands and, for the sake of exposition, we can
organize them by scope.
The scope of the **general commands** is the entire Barman server,
that can backup many PostgreSQL servers. **Server commands**, instead,
act only on a specified server. **Backup commands** work on a backup,
which is taken from a certain server.
The following list includes the general commands.
## `cron`
You can perform maintenance operations, on both WAL files and backups,
using the `cron` command:
``` bash
barman cron
```
> **NOTE:**
> This command should be executed in a _cron script_. Our
> recommendation is to schedule `barman cron` to run every minute. If
> you installed Barman using the rpm or debian package, a cron entry
> running on every minute will be created for you.
`barman cron` executes WAL archiving operations concurrently on a
server basis, and this also enforces retention policies on those
servers that have:
- `retention_policy` not empty and valid;
- `retention_policy_mode` set to `auto`.
The `cron` command ensures that WAL streaming is started for those
servers that have requested it, by transparently executing the
`receive-wal` command.
## `diagnose`
The `diagnose` command creates a JSON report useful for diagnostic and
support purposes. This report contains information for all configured
servers.
> **IMPORTANT:**
> Even if the diagnose is written in JSON and that format is thought
> to be machine readable, its structure is not to be considered part
> of the interface. Format can change between different Barman versions.
## `list-server`
You can display the list of active servers that have been configured
for your backup system with:
``` bash
barman list-server
```
A machine readble output can be obtained with the `--minimal` option:
``` bash
barman list-server --minimal
```
barman-2.1/doc/manual/24-wal_archiving.en.md 0000664 0001750 0001750 00000003711 13025741466 021231 0 ustar gcalacoci gcalacoci ## WAL archiving via `archive_command`
The `archive_command` is the traditional method to archive WAL files.
The value of this PostgreSQL configuration parameter must be a shell
command to be executed by the PostgreSQL server to copy the WAL files
to the Barman incoming directory.
You can retrieve the incoming WALs directory using the `show-server`
Barman command and looking for the `incoming_wals_directory` value:
``` bash
barman@backup$ barman show-server pg |grep incoming_wals_directory
incoming_wals_directory: /var/lib/barman/pg/incoming
```
> **IMPORTANT:**
> PostgreSQL 9.5 introduced support for WAL file archiving using
> `archive_command` from a standby. This feature is not yet implemented
> in Barman.
Edit the `postgresql.conf` file of the PostgreSQL instance on the `pg`
database and activate the archive mode:
``` ini
archive_mode = on
wal_level = 'replica'
archive_command = 'rsync -a %p barman@backup:INCOMING_WALS_DIRECTORY/%f'
```
Make sure you change the `INCOMING_WALS_DIRECTORY` placeholder with
the value returned by the `barman show-server pg` command above.
Restart the PostgreSQL server.
In order to test that continuous archiving is on and properly working,
you need to check both the PostgreSQL server and the backup server. In
particular, you need to check that WAL files are correctly collected
in the destination directory.
## Verification of WAL archiving configuration
In order to improve the verification of the WAL archiving process, the
`switch-xlog` command has been developed:
``` bash
barman@backup$ barman switch-xlog --force --archive pg
```
The above command will force PostgreSQL to switch WAL file and
trigger the archiving process in Barman. Barman will wait for one
file to arrive within 30 seconds (you can change the timeout through
the `--archive-timeout` option). If no WAL file is received, an error
is returned.
You can verify if the WAL archiving has been correctly configured using
the `barman check` command.
barman-2.1/doc/barman.1 0000664 0001750 0001750 00000036227 13033425212 015306 0 ustar gcalacoci gcalacoci .TH "BARMAN" "1" "January 5, 2017" "Barman User manuals" "Version 2.1"
.SH NAME
.PP
barman \- Backup and Recovery Manager for PostgreSQL
.SH SYNOPSIS
.PP
barman [\f[I]OPTIONS\f[]] \f[I]COMMAND\f[]
.SH DESCRIPTION
.PP
Barman is an administration tool for disaster recovery of PostgreSQL
servers written in Python and maintained by 2ndQuadrant.
Barman can perform remote backups of multiple servers in business
critical environments and helps DBAs during the recovery phase.
.SH OPTIONS
.TP
.B \-v, \-\-version
Show program version number and exit.
.RS
.RE
.TP
.B \-q, \-\-quiet
Do not output anything.
Useful for cron scripts.
.RS
.RE
.TP
.B \-h, \-\-help
Show a help message and exit.
.RS
.RE
.TP
.B \-c \f[I]CONFIG\f[], \-\-config \f[I]CONFIG\f[]
Use the specified configuration file.
.RS
.RE
.SH COMMANDS
.PP
Important: every command has a help option
.TP
.B archive\-wal \f[I]SERVER_NAME\f[]
Get any incoming xlog file (both through standard
\f[C]archive_command\f[] and streaming replication, where applicable)
and moves them in the WAL archive for that server.
If necessary, apply compression when requested by the user.
.RS
.RE
.TP
.B cron
Perform maintenance tasks, such as enforcing retention policies or WAL
files management.
.RS
.RE
.TP
.B list\-server
Show all the configured servers, and their descriptions.
.RS
.RE
.TP
.B show\-server \f[I]SERVER_NAME\f[]
Show information about \f[C]SERVER_NAME\f[], including:
\f[C]conninfo\f[], \f[C]backup_directory\f[], \f[C]wals_directory\f[]
and many more.
Specify \f[C]all\f[] as \f[C]SERVER_NAME\f[] to show information about
all the configured servers.
.RS
.RE
.TP
.B status \f[I]SERVER_NAME\f[]
Show information about the status of a server, including: number of
available backups, \f[C]archive_command\f[], \f[C]archive_status\f[] and
many more.
For example:
.RS
.RE
.IP
.nf
\f[C]
Server\ quagmire:
\ \ Description:\ The\ Giggity\ database
\ \ Passive\ node:\ False
\ \ PostgreSQL\ version:\ 9.3.9
\ \ pgespresso\ extension:\ Not\ available
\ \ PostgreSQL\ Data\ directory:\ /srv/postgresql/9.3/data
\ \ PostgreSQL\ \[aq]archive_command\[aq]\ setting:\ rsync\ \-a\ %p\ barman\@backup:/var/lib/barman/quagmire/incoming
\ \ Last\ archived\ WAL:\ 0000000100003103000000AD
\ \ Current\ WAL\ segment:\ 0000000100003103000000AE
\ \ Retention\ policies:\ enforced\ (mode:\ auto,\ retention:\ REDUNDANCY\ 2,\ WAL\ retention:\ MAIN)
\ \ No.\ of\ available\ backups:\ 2
\ \ First\ available\ backup:\ 20150908T003001
\ \ Last\ available\ backup:\ 20150909T003001
\ \ Minimum\ redundancy\ requirements:\ satisfied\ (2/1)
\f[]
.fi
.TP
.B check \f[I]SERVER_NAME\f[]
Show diagnostic information about \f[C]SERVER_NAME\f[], including: Ssh
connection check, PostgreSQL version, configuration and backup
directories, archiving process, streaming process, replication slots,
etc.
Specify \f[C]all\f[] as \f[C]SERVER_NAME\f[] to show diagnostic
information about all the configured servers.
.RS
.TP
.B \-\-nagios
Nagios plugin compatible output
.RS
.RE
.RE
.TP
.B diagnose
Collect diagnostic information about the server where barman is
installed and all the configured servers, including: global
configuration, SSH version, Python version, \f[C]rsync\f[] version, as
well as current configuration and status of all servers.
.RS
.RE
.TP
.B backup \f[I]SERVER_NAME\f[]
Perform a backup of \f[C]SERVER_NAME\f[] using parameters specified in
the configuration file.
Specify \f[C]all\f[] as \f[C]SERVER_NAME\f[] to perform a backup of all
the configured servers.
.RS
.TP
.B \-\-immediate\-checkpoint
forces the initial checkpoint to be done as quickly as possible.
Overrides value of the parameter \f[C]immediate_checkpoint\f[], if
present in the configuration file.
.RS
.RE
.TP
.B \-\-no\-immediate\-checkpoint
forces to wait for the checkpoint.
Overrides value of the parameter \f[C]immediate_checkpoint\f[], if
present in the configuration file.
.RS
.RE
.TP
.B \-\-reuse\-backup [INCREMENTAL_TYPE]
Overrides \f[C]reuse_backup\f[] option behaviour.
Possible values for \f[C]INCREMENTAL_TYPE\f[] are:
.RS
.IP \[bu] 2
\f[I]off\f[]: do not reuse the last available backup;
.IP \[bu] 2
\f[I]copy\f[]: reuse the last available backup for a server and create a
copy of the unchanged files (reduce backup time);
.IP \[bu] 2
\f[I]link\f[]: reuse the last available backup for a server and create a
hard link of the unchanged files (reduce backup time and space);
.PP
\f[C]link\f[] is the default target if \f[C]\-\-reuse\-backup\f[] is
used and \f[C]INCREMENTAL_TYPE\f[] is not explicited.
.RE
.TP
.B \-\-retry\-times
Number of retries of base backup copy, after an error.
Used during both backup and recovery operations.
Overrides value of the parameter \f[C]basebackup_retry_times\f[], if
present in the configuration file.
.RS
.RE
.TP
.B \-\-no\-retry
Same as \f[C]\-\-retry\-times\ 0\f[]
.RS
.RE
.TP
.B \-\-retry\-sleep
Number of seconds of wait after a failed copy, before retrying.
Used during both backup and recovery operations.
Overrides value of the parameter \f[C]basebackup_retry_sleep\f[], if
present in the configuration file.
.RS
.RE
.RE
.TP
.B list\-backup \f[I]SERVER_NAME\f[]
Show available backups for \f[C]SERVER_NAME\f[].
This command is useful to retrieve a backup ID.
For example:
.RS
.RE
.IP
.nf
\f[C]
servername\ 20111104T102647\ \-\ Fri\ Nov\ \ 4\ 10:26:48\ 2011\ \-\ Size:\ 17.0\ MiB\ \-\ WAL\ Size:\ 100\ B
\f[]
.fi
.IP
.nf
\f[C]
In\ this\ case,\ *20111104T102647*\ is\ the\ backup\ ID.
\f[]
.fi
.TP
.B show\-backup \f[I]SERVER_NAME\f[] \f[I]BACKUP_ID\f[]
Show detailed information about a particular backup, identified by the
server name and the backup ID.
See the Backup ID shortcuts (#shortcuts) section below for available
shortcuts.
For example:
.RS
.RE
.IP
.nf
\f[C]
Backup\ 20150828T130001:
\ \ Server\ Name\ \ \ \ \ \ \ \ \ \ \ \ :\ quagmire
\ \ Status\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ :\ DONE
\ \ PostgreSQL\ Version\ \ \ \ \ :\ 90402
\ \ PGDATA\ directory\ \ \ \ \ \ \ :\ /srv/postgresql/9.4/main/data
\ \ Base\ backup\ information:
\ \ \ \ Disk\ usage\ \ \ \ \ \ \ \ \ \ \ :\ 12.4\ TiB\ (12.4\ TiB\ with\ WALs)
\ \ \ \ Incremental\ size\ \ \ \ \ :\ 4.9\ TiB\ (\-60.02%)
\ \ \ \ Timeline\ \ \ \ \ \ \ \ \ \ \ \ \ :\ 1
\ \ \ \ Begin\ WAL\ \ \ \ \ \ \ \ \ \ \ \ :\ 0000000100000CFD000000AD
\ \ \ \ End\ WAL\ \ \ \ \ \ \ \ \ \ \ \ \ \ :\ 0000000100000D0D00000008
\ \ \ \ WAL\ number\ \ \ \ \ \ \ \ \ \ \ :\ 3932
\ \ \ \ WAL\ compression\ ratio:\ 79.51%
\ \ \ \ Begin\ time\ \ \ \ \ \ \ \ \ \ \ :\ 2015\-08\-28\ 13:00:01.633925+00:00
\ \ \ \ End\ time\ \ \ \ \ \ \ \ \ \ \ \ \ :\ 2015\-08\-29\ 10:27:06.522846+00:00
\ \ \ \ Begin\ Offset\ \ \ \ \ \ \ \ \ :\ 1575048
\ \ \ \ End\ Offset\ \ \ \ \ \ \ \ \ \ \ :\ 13853016
\ \ \ \ Begin\ XLOG\ \ \ \ \ \ \ \ \ \ \ :\ CFD/AD180888
\ \ \ \ End\ XLOG\ \ \ \ \ \ \ \ \ \ \ \ \ :\ D0D/8D36158
\ \ WAL\ information:
\ \ \ \ No\ of\ files\ \ \ \ \ \ \ \ \ \ :\ 35039
\ \ \ \ Disk\ usage\ \ \ \ \ \ \ \ \ \ \ :\ 121.5\ GiB
\ \ \ \ WAL\ rate\ \ \ \ \ \ \ \ \ \ \ \ \ :\ 275.50/hour
\ \ \ \ Compression\ ratio\ \ \ \ :\ 77.81%
\ \ \ \ Last\ available\ \ \ \ \ \ \ :\ 0000000100000D95000000E7
\ \ Catalog\ information:
\ \ \ \ Retention\ Policy\ \ \ \ \ :\ not\ enforced
\ \ \ \ Previous\ Backup\ \ \ \ \ \ :\ 20150821T130001
\ \ \ \ Next\ Backup\ \ \ \ \ \ \ \ \ \ :\ \-\ (this\ is\ the\ latest\ base\ backup)
\f[]
.fi
.TP
.B list\-files \f[I][OPTIONS]\f[] \f[I]SERVER_NAME\f[]
\f[I]BACKUP_ID\f[]
List all the files in a particular backup, identified by the server name
and the backup ID.
See the Backup ID shortcuts (#shortcuts) section below for available
shortcuts.
.RS
.TP
.B \-\-target \f[I]TARGET_TYPE\f[]
Possible values for TARGET_TYPE are:
.RS
.IP \[bu] 2
\f[I]data\f[]: lists just the data files;
.IP \[bu] 2
\f[I]standalone\f[]: lists the base backup files, including required WAL
files;
.IP \[bu] 2
\f[I]wal\f[]: lists all the WAL files between the start of the base
backup and the end of the log / the start of the following base backup
(depending on whether the specified base backup is the most recent one
available);
.IP \[bu] 2
\f[I]full\f[]: same as data + wal.
.PP
The default value is \f[C]standalone\f[].
.RE
.RE
.TP
.B rebuild\-xlogdb \f[I]SERVER_NAME\f[]
Perform a rebuild of the WAL file metadata for \f[C]SERVER_NAME\f[] (or
every server, using the \f[C]all\f[] shortcut) guessing it from the disk
content.
The metadata of the WAL archive is contained in the \f[C]xlog.db\f[]
file, and every Barman server has its own copy.
.RS
.RE
.TP
.B recover \f[I][OPTIONS]\f[] \f[I]SERVER_NAME\f[] \f[I]BACKUP_ID\f[]
\f[I]DESTINATION_DIRECTORY\f[]
Recover a backup in a given directory (local or remote, depending on the
\f[C]\-\-remote\-ssh\-command\f[] option settings).
See the Backup ID shortcuts (#shortcuts) section below for available
shortcuts.
.RS
.TP
.B \-\-target\-tli \f[I]TARGET_TLI\f[]
Recover the specified timeline.
.RS
.RE
.TP
.B \-\-target\-time \f[I]TARGET_TIME\f[]
Recover to the specified time.
.RS
.PP
You can use any valid unambiguous representation (e.g: "YYYY\-MM\-DD
HH:MM:SS.mmm").
.RE
.TP
.B \-\-target\-xid \f[I]TARGET_XID\f[]
Recover to the specified transaction ID.
.RS
.RE
.TP
.B \-\-target\-name \f[I]TARGET_NAME\f[]
Recover to the named restore point previously created with the
\f[C]pg_create_restore_point(name)\f[] (for PostgreSQL 9.1 and above
users).
.RS
.RE
.TP
.B \-\-exclusive
Set target xid to be non inclusive.
.RS
.RE
.TP
.B \-\-tablespace \f[I]NAME:LOCATION\f[]
Specify tablespace relocation rule.
.RS
.RE
.TP
.B \-\-remote\-ssh\-command \f[I]SSH_COMMAND\f[]
This options activates remote recovery, by specifying the secure shell
command to be launched on a remote host.
This is the equivalent of the "ssh_command" server option in the
configuration file for remote recovery.
Example: \[aq]ssh postgres\@db2\[aq].
.RS
.RE
.TP
.B \-\-retry\-times \f[I]RETRY_TIMES\f[]
Number of retries of data copy during base backup after an error.
Overrides value of the parameter \f[C]basebackup_retry_times\f[], if
present in the configuration file.
.RS
.RE
.TP
.B \-\-no\-retry
Same as \f[C]\-\-retry\-times\ 0\f[]
.RS
.RE
.TP
.B \-\-retry\-sleep
Number of seconds of wait after a failed copy, before retrying.
Overrides value of the parameter \f[C]basebackup_retry_sleep\f[], if
present in the configuration file.
.RS
.RE
.RE
.TP
.B get\-wal \f[I][OPTIONS]\f[] \f[I]SERVER_NAME\f[] \f[I]WAL_ID\f[]
Retrieve a WAL file from the \f[C]xlog\f[] archive of a given server.
By default, the requested WAL file, if found, is returned as
uncompressed content to \f[C]STDOUT\f[].
The following options allow users to change this behaviour:
.RS
.TP
.B \-o \f[I]OUTPUT_DIRECTORY\f[]
destination directory where the \f[C]get\-wal\f[] will deposit the
requested WAL
.RS
.RE
.TP
.B \-z
output will be compressed using gzip
.RS
.RE
.TP
.B \-j
output will be compressed using bzip2
.RS
.RE
.TP
.B \-p \f[I]SIZE\f[]
peek from the WAL archive up to \f[I]SIZE\f[] WAL files, starting from
the requested one.
\[aq]SIZE\[aq] must be an integer >= 1.
When invoked with this option, get\-wal returns a list of zero to
\[aq]SIZE\[aq] WAL segment names, one per row.
.RS
.RE
.RE
.TP
.B switch\-xlog \f[I]SERVER_NAME\f[]
Execute pg_switch_xlog() on the target server
.RS
.TP
.B \-\-force
Forces the switch by executing CHECKPOINT before pg_switch_xlog().
\f[I]IMPORTANT:\f[] executing a CHECKPOINT might increase I/O load on a
PostgreSQL server.
Use this option with care.
.RS
.RE
.TP
.B \-\-archive
Wait for one xlog file to be archived.
If after a defined amount of time (default: 30 seconds) no xlog file is
archived, Barman will teminate with failure exit code.
.RS
.RE
.TP
.B \-\-archive\-timeout \f[I]TIMEOUT\f[]
Specifies the amount of time in seconds (default: 30 seconds) the
archiver will wait for a new xlog file to be archived before timing out.
.RS
.RE
.RE
.TP
.B receive\-wal \f[I]SERVER_NAME\f[]
Start the stream of transaction logs for a server.
The process relies on \f[C]pg_receivexlog\f[] to receive WAL files from
the PostgreSQL servers through the streaming protocol.
.RS
.TP
.B \-\-stop
stop the receive\-wal process for the server
.RS
.RE
.TP
.B \-\-reset
reset the status of receive\-wal, restarting the streaming from the
current WAL file of the server
.RS
.RE
.TP
.B \-\-create\-slot
create the physical replication slot configured with the
\f[C]slot_name\f[] configuration parameter
.RS
.RE
.TP
.B \-\-drop\-slot
drop the physical replication slot configured with the
\f[C]slot_name\f[] configuration parameter
.RS
.RE
.RE
.TP
.B delete \f[I]SERVER_NAME\f[] \f[I]BACKUP_ID\f[]
Delete the specified backup.
Backup ID shortcuts (#shortcuts) section below for available shortcuts.
.RS
.RE
.TP
.B replication\-status \f[I][OPTIONS]\f[] \f[I]SERVER_NAME\f[]
Shows live information and status of any streaming client attached to
the given server (or servers).
Default behaviour can be changed through the following options:
.RS
.TP
.B \-\-minimal
machine readable output (default: False)
.RS
.RE
.TP
.B \-\-target \f[I]TARGET_TYPE\f[]
Possible values for TARGET_TYPE are:
.RS
.IP \[bu] 2
\f[I]hot\-standby\f[]: lists only hot standby servers
.IP \[bu] 2
\f[I]wal\-streamer\f[]: lists only WAL streaming clients, such as
pg_receivexlog
.IP \[bu] 2
\f[I]all\f[]: any streaming client (default)
.RE
.RE
.SH BACKUP ID SHORTCUTS
.PP
Rather than using the timestamp backup ID, you can use any of the
following shortcuts/aliases to identity a backup for a given server:
.TP
.B first
Oldest available backup for that server, in chronological order.
.RS
.RE
.TP
.B last
Latest available backup for that server, in chronological order.
.RS
.RE
.TP
.B latest
same ast \f[I]last\f[].
.RS
.RE
.TP
.B oldest
same ast \f[I]first\f[].
.RS
.RE
.SH EXIT STATUS
.TP
.B 0
Success
.RS
.RE
.TP
.B Not zero
Failure
.RS
.RE
.SH SEE ALSO
.PP
\f[C]barman\f[] (5).
.SH BUGS
.PP
Barman has been extensively tested, and is currently being used in
several production environments.
However, we cannot exclude the presence of bugs.
.PP
Any bug can be reported via the Sourceforge bug tracker.
Along the bug submission, users can provide developers with diagnostics
information obtained through the \f[C]barman\ diagnose\f[] command.
.SH AUTHORS
.PP
In alphabetical order:
.IP \[bu] 2
Gabriele Bartolini (project leader)
.IP \[bu] 2
Jonathan Battiato (QA/testing)
.IP \[bu] 2
Stefano Bianucci (developer, intern from University of Florence)
.IP \[bu] 2
Giuseppe Broccolo (QA/testing)
.IP \[bu] 2
Giulio Calacoci (developer)
.IP \[bu] 2
Francesco Canovai (QA/testing)
.IP \[bu] 2
Leonardo Cecchi (developer)
.IP \[bu] 2
Gianni Ciolli (QA/testing)
.IP \[bu] 2
Britt Cole (documentation)
.IP \[bu] 2
Marco Nenciarini (lead developer)
.IP \[bu] 2
Rubens Souza (QA/testing)
.PP
Past contributors:
.IP \[bu] 2
Carlo Ascani
.SH RESOURCES
.IP \[bu] 2
Homepage:
.IP \[bu] 2
Documentation:
.IP \[bu] 2
Professional support:
.SH COPYING
.PP
Barman is the property of 2ndQuadrant Italia and its code is distributed
under GNU General Public License v3.
.PP
Copyright (C) 2011\-2016 2ndQuadrant Italia Srl \-
.
.SH AUTHORS
2ndQuadrant Italy .
barman-2.1/ChangeLog 0000664 0001750 0001750 00000352366 13033425616 015006 0 ustar gcalacoci gcalacoci 2017-01-05 Giulio Calacoci
Update the ChangeLog file
Set version to 2.1
2016-12-22 Marco Nenciarini
Avoid importing barman.utils module twice
2017-01-03 Gabriele Bartolini
Make retry hook script options global (bugfix)
2016-12-27 Marco Nenciarini
Update the ChangeLog file
2016-12-26 Gabriele Bartolini
Release 2.1a1
2016-12-21 Marco Nenciarini
Add PostgreSQL 10 versioning scheme compatibility
Fix #73
Exclude extraneous files during tablespace copy
PostgreSQL stores the tablespace data in a directory defined in the
PostgreSQL source as:
"PG_" PG_MAJORVERSION "_" CATALOG_VERSION_NO
The CATALOG_VERSION_NO is not easy to retrieve, so we copy
"PG_" + PG_MAJORVERSION + "_*"
It could select some spurious directory if a development or a beta
version have been used, but it's good enough for a production system
as it filters out other major versions.
Fix #74
2016-12-20 Marco Nenciarini
Align the list of excluded files to pg_basebackup
Fixes #72
2016-12-18 Marco Nenciarini
Return failure when 'barman get-wal' command is used on inactive servers
The 'barman get-wal' command is meant to be used used as PostgreSQL's
restore_command, so it must succeed only if the target WAL file has
been actually returned.
2016-12-17 Marco Nenciarini
Add last archived WAL info to diagnose output
Show in diagnose the last archived WAL per timeline
switch-xlog --archive waits for the closed WAL file
Modify the switch-xlog --archive behaviour to wait for the specific WAL
file, instead of a generic one. This is faster and it avoids possible
issues when a higher timeline is present in the archive.
2016-12-03 Marco Nenciarini
Add --archive and --archive-timeout options to switch-xlog command
Improve the switch-xlog command adding the --archive and
--archive-timeout options.
The --archive option will start an archiver process for the selected
server, and will wait for 30 seconds for a new xlog to be archived.
The --archive-timeout TIMEOUT will allow the user to increase or
decrease the number of seconds the archiver process will wait
for the archival of a new xlog file, before exiting.
2016-11-26 Gabriele Bartolini
Make streaming application name options global
Make both 'streaming_archiver_name' and 'streaming_backup_name'
global options, as per documentation. Bugfix.
Fixes #57
2016-11-27 Marco Nenciarini
Fix rsync failures due to files truncated during transfer
Fixes #64
2016-11-26 Gabriele Bartolini
Fix typo in documentation
2016-11-17 Gabriele Bartolini
Convert README in README.rst
Suggested by ckeeney. Closes #69
2016-11-24 Giulio Calacoci
Correctly handle compressed history files
Fix how barman behave when a compressed history files is found.
Fixes: #66
Avoid dereferencing symlinks in pg_tblspc when preparing recovery
This commit fixes a bug triggered by the presence of a broken tablespace
symlink in the target pg_tblspc directory when recovery starts.
Fixes: #55
2016-10-06 Martin Swiech
Fix comparation of last archiving failure
Compare the time of last failure instead of the WAL name to detect
archiving failures.
Closes: #40, #58
2016-11-13 Marco Nenciarini
Avoid failing recovery if postgresql.conf is not writable
Fixes: #68
2016-11-08 Giulio Calacoci
Add start time and execution time to delete output
Improve the output of `barman delete` command by adding some execution
time statistics.
Additionally, the utility function that prints human readable time deltas
has been improved to correctly handle a wider range of values.
Closes #67
2016-10-26 Ian Barwick
Change references from non-existent `path` option to `path_prefix`
2016-10-28 Marco Nenciarini
Exclude files within pg_replslot from backups
Thanks to Magnus Hagander for reporting the issue.
Ref: https://www.postgresql.org/docs/current/static/continuous-archiving.html#BACKUP-LOWLEVEL-BASE-BACKUP-DATA
Fixes #65
2016-10-20 Giulio Calacoci
Workaroud to setuptools testrunner bug.
Ref links:
* https://bitbucket.org/pypa/setuptools/issue/196
* https://github.com/pypa/setuptools/issues/196
2016-10-17 Giulio Calacoci
Fix output of the replication-status command.
The replication-status command used to emit an exception in some
PostgreSQL releases, due to the different structure of the
`pg_stat_replication` view.
This patch fixes the issue checking for the existence of the fields
before writing them to the output.
Fixes #56
2016-10-11 Marco Nenciarini
Move isort configuration in setup.cfg
2016-10-11 Gabriele Bartolini
Add a note in docs about concurrent backup for 9.6
2016-09-30 Giulio Calacoci
Split the logic of the copy method in CopyController
This patch creates an analisys phase and a copy phase. the analisys
phase is responsible for the preparation of the list of files that we
want to copy. The copy phase is responsible for the execution of the
actual copy using rsync.
2016-10-02 Gabriele Bartolini
Fix level of section about WAL verification
2016-09-27 Giulio Calacoci
Update sphinx index page
2016-09-26 Gabriele Bartolini
Set version 2.1a1
2016-09-26 Marco Nenciarini
Update the ChangeLog file
2016-09-26 Gabriele Bartolini
Update architecture diagrams
2016-09-26 Marco Nenciarini
Review of example configurations in doc/barman.d/
2016-09-26 Leonardo Cecchi
Fix image size in manual
2016-09-26 Gabriele Bartolini
Set version to 2.0
2016-09-25 Gabriele Bartolini
General review of the manual before 2.0 release
2016-09-23 Giulio Calacoci
Add section `Upgrade from Barman 1.X` to manual.
2016-09-23 Leonardo Cecchi
Improve documentation of get-wal command
This patch improves the documentation of the get-wal command for the
local recovery, specifying that the command must be executed as `barman`
user and not as the user which is running the PostgreSQL server.
It also includes details about the barman-wal-restore command.
2016-09-23 Gabriele Bartolini
Update INSTALL, README and TODO files
2016-09-23 Leonardo Cecchi
Add Windows server documentation
Add a Windows server section in the "Setup a new server in Barman"
chapter.
2016-09-21 Leonardo Cecchi
Fix wrongly named max_replication_slots configuration parameter
2016-09-20 Marco Nenciarini
Update the ChangeLog file
2016-09-19 csawyerYumaed
Improve barman(5) manual page, adding conninfo
Add additional information to documentation around conninfo.
Closes #53
2016-09-20 Marco Nenciarini
Set version to 2.0b1
Remove the quickstart guide
At the moment, it is an empty document. It may be reintroduced later.
2016-09-20 Leonardo Cecchi
Improve configurantion error message
The configuration error message raised when a parameter has an invalid
value was not telling the wrong value but only the key.
This patch improves the error message in that case.
2016-09-16 Giulio Calacoci
Add troubleshooting section to manual
2016-09-16 Leonardo Cecchi
Improve "Features in detail" section
2016-09-19 Giulio Calacoci
Avoid logging tracebacks in case of expected failures
There was a few places in the code where an expected error was producing
a traceback in the logs. They have been replaced with a simple error
output.
2016-09-16 Giulio Calacoci
Integrate barman-wal-restore in barman recovery
During a remote restore, use the barman-wal-restore command instead of
calling `barman get-wal` on the remote machine. It requires the
barman-cli package to be installed on the target server.
Force special tox environments to use a specific Python version
Use Python 2.7 in the 'flake8' environment, because it doesn't work with
Python 2.6.
Use Python 2.6 in the 'minimal' envorinment, because it doesn't work
with Python 3.x due to the presence of Python-2-only modules.
Enable the minimal target in Travis-CI again, as it was failing due to
the default Python version used there (Python 3.5).
2016-09-19 Marco Nenciarini
Fix wrongly named database parameter in configuration examples
Thanks to Abhijit Menon-Sen who found it.
2016-09-14 Leonardo Cecchi
Add "server commands" documentation section
This patch complete the section "Server commands" inside the Barman manual
2016-09-14 Giulio Calacoci
Backup command section of the manual
2016-09-14 Leonardo Cecchi
Add the "general commands" documentation section
This patch complete the section "general commands" inside the Barman
manual.
2016-09-13 Leonardo Cecchi
Add the "setup new server" documentation section
This patch complete the work on the "Setup of a new server in Barman"
documentation section.
2016-09-13 Giulio Calacoci
Make sure 'dbname' parameter is not set for streaming connections
Also fix an issue in the parse_dsn method, many thanks to
Théophile Helleboid for reporting it.
2016-09-15 Marco Nenciarini
Fix flake8 plugin versions
2016-09-13 Marco Nenciarini
Make sure path_prefix is honored in every Command subclass
Commit ad8002745326fe20a2c4ce2a5e9115a26c5a5f49 contains a typo
so it was not working at all. The typo is fixed in this commit.
2016-09-13 Giulio Calacoci
Pass `--no-password` option to pg_basebackup and pg_receivexlog
Barman will never send any password through the pg_basebackup
or
pg_receivexlog standard input channel. Invoking them with the
`--no-password` option will produce a cleaner error message in case of
authentication failure.
Modify tests accordingly.
2016-09-12 Giulio Calacoci
Improve the warning message for configuration files outside PGDATA
Now the `backup_method=postgres` code emits the warning message for
configuration files that are located outside PGDATA only when that's
the actual case.
2016-09-12 Gabriele Bartolini
Moved barman-wal-restore in barman-cli project
2016-09-12 Marco Nenciarini
Enable `show-server` command on inactive or disabled servers
2016-09-09 Marco Nenciarini
Add missing release date on manual front page
Disable minimal tox env on travis-ci as it doesn't work on py35
Update the ChangeLog file
Prepare release 2.0a1
Run minimal and flake8 tox envs in travis-ci
2016-09-09 Gabriele Bartolini
Fix flake8 error
2016-08-26 Gabriele Bartolini
First stub of documentation for 2.0
This is a first attempt to change the documentation in Barman,
which creates two artifacts: a manual and a quickstart guide.
The work is currently incomplete, but it is ok for a first
alpha release, given that all required information from a
technical point of view is in man pages.
The reason for this is that now Barman does not have a sequential
installation procedure, but different options for backup and
WAL archiving that produce several possibile combinations of
setups.
2016-09-09 Marco Nenciarini
Set last supported version by argh for Python 2.x
Add minimal checks in Tox
2016-09-09 Gabriele Bartolini
Release notes for 2.0a1
2016-09-09 Marco Nenciarini
Rename barman-wal-restore.py to barman-wal-restore
2016-09-09 Leonardo Cecchi
Fix exception in barman check
Barman check raise an exception when streaming_archiver is on and
streaming_conninfo is not configured.
This patch controls if the streaming connection is correctly configured
before checking the pg_basebackup compatibility, removing the exception
cause.
2016-09-09 Marco Nenciarini
Remove unit test dependency on environment
Previously unit tests were failing the $USER environment variable was
not defined. This patch removes this requirement.
2016-09-08 Leonardo Cecchi
Make sure path is used in every Command class
Switch xlog after a concurrent backup
For a backup to be usable, all the XLOG files written between the backup
start and the backup end must be archived.
This patch makes PostgreSQL switch the current XLOG file after a
concurrent backup either using rsync or streaming backup mode, in this
way the WAL file containing the backup end position will be archived as
soon as possible.
2016-09-07 Giulio Calacoci
Add retry support when `backup_method=postgres`
Modify the backup executor to handle the retry of a streaming
backup with pg_basebackup (`backup_method=postgres`).
Between every retry attempt, the PGDATA and the directories that
contain the tablespaces are removed. Therefore, subsequent
backup attempts restart from scratch.
2016-09-07 Gabriele Bartolini
Fix E303 too many blank lines (flake8)
2016-09-07 Marco Nenciarini
Support /etc/barman.d in rpm packaging
2016-09-01 Marco Nenciarini
Redesign retry logic for backup and recovery
The CommandWrapper class is now responsible to specifically
design the logic of a retry operation in case of failure.
Previously, retry was managed at an outer level, by the
backup manager and the executor classes.
With this new approach, every single command can now develop its
own retry logic.
2016-09-05 Giulio Calacoci
Force `archiver=on` when no archiver is active.
This is a back-compatibility feature that will be kept
in Barman for a few releases, and mitigate the upgrade
to version 2.0 (which requires explicit setting of
an archiving method - before, `archiver` was `on` by default)
2016-09-05 Gabriele Bartolini