barman.xlog module#

This module contains functions to retrieve information about xlog files

barman.xlog.DEFAULT_XLOG_SEG_SIZE = 16777216#: XLOG_SEG_SIZE is the size of a single WAL file. This must be a power of 2 and larger than XLOG_BLCKSZ (preferably, a great deal larger than XLOG_BLCKSZ).

class barman.xlog.HistoryFileData(tli, parent_tli, switchpoint, reason)#

Bases: tuple

This namedtuple is a container for the information contained inside history files

_asdict()#: Return a new dict which maps field names to their values.

_field_defaults = {}#

_fields = ('tli', 'parent_tli', 'switchpoint', 'reason')#

classmethod _make(iterable)#: Make a new HistoryFileData object from a sequence or iterable

_replace(**kwds)#: Return a new HistoryFileData object replacing specified fields with new values

parent_tli#: Alias for field number 1

reason#: Alias for field number 3

switchpoint#: Alias for field number 2

tli#: Alias for field number 0

barman.xlog._validate_timeline(timeline)View on GitHub #: Check that timeline is a valid timeline value.

barman.xlog._wal_archive_filter_fun(timeline, wal)View on GitHub #

barman.xlog.check_archive_usable(existing_wals, timeline=None)View on GitHub #: Carry out pre-flight checks on the existing content of a WAL archive to determine if it is safe to archive WALs from the supplied timeline.

barman.xlog.decode_hash_dir(hash_dir)View on GitHub #

Get the timeline and log from a hash dir prefix.

Parameters:: hash_dir (str) – A string representing the prefix used when determining the folder or object key prefix under which Barman will store a given WAL segment. This prefix is composed of the timeline and the higher 32-bit number of the WAL segment.
Return type:: List[int]
Returns:: A list of two elements where the first item is the timeline and the second is the higher 32-bit number of the WAL segment.

barman.xlog.decode_history_file(wal_info, comp_manager)View on GitHub #

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.

Parameters:

wal_info (barman.infofile.WalFileInfo) – history file obj
comp_manager – compression manager used in case of history file compression

Return List[HistoryFileData]:

information from the history file

barman.xlog.decode_segment_name(path)View on GitHub #

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.

Parameters:: path (str) – the file name to decode
Return type:: list[int]

barman.xlog.diff_lsn(lsn_string1, lsn_string2)View on GitHub #

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.

Parameters:

lsn_string1 (str) – the string XLOG location, i.e. ‘2/82000168’
lsn_string2 (str) – the string XLOG location, i.e. ‘2/82000168’

Return type:

int

barman.xlog.encode_history_file_name(tli)View on GitHub #

Build the history file name based on timeline

Return str:: history file name

barman.xlog.encode_segment_name(tli, log, seg)View on GitHub #

Build the xlog segment name based on timeline, log ID and segment ID

Parameters:

tli (int) – timeline number
log (int) – log number
seg (int) – segment number

Return str:

segment file name

barman.xlog.format_lsn(lsn)View on GitHub #

Transform a numeric XLOG location, in the corresponding %X/%X string representation

Parameters:: lsn (int) – numeric XLOG location
Return type:: str

barman.xlog.generate_segment_names(begin, end=None, version=None, xlog_segment_size=None)View on GitHub #

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.

If the XLOG segment size is known, this generator is precise, switching to the next file when required.

It the XLOG segment size is unknown, this generator will generate all the possible XLOG file names. The size of an XLOG segment can be every power of 2 between the XLOG block size (8Kib) and the size of a log segment (4Gib)

Parameters:

begin (str) – begin segment name
end (str|None) – optional end segment name
version (int|None) – optional postgres version as an integer (e.g. 90301 for 9.3.1)
xlog_segment_size (int) – the size of a XLOG segment

Return type:

collections.Iterable[str]

Raise:

BadXlogSegmentName

barman.xlog.hash_dir(path)View on GitHub #

Get the directory where the xlog segment will be stored

It can handle either a full file path or a simple file name.

Parameters:: path (str|unicode) – xlog file name
Return str:: directory name

barman.xlog.is_any_xlog_file(path)View on GitHub #

Return True if the xlog is either a WAL segment, a .backup file or a .history file, False otherwise.