1
0
mirror of https://github.com/pgbackrest/pgbackrest.git synced 2025-12-23 23:51:07 +02:00
Files
pgbackrest/lib/BackRest/Backup.pm

1046 lines
42 KiB
Perl
Raw Normal View History

v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
####################################################################################################################################
# BACKUP MODULE
####################################################################################################################################
package BackRest::Backup;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
use threads;
use strict;
use warnings FATAL => qw(all);
use Carp qw(confess);
use Exporter qw(import);
use Fcntl 'SEEK_CUR';
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
use File::Basename;
use File::Path qw(remove_tree);
use Scalar::Util qw(looks_like_number);
use Thread::Queue;
use lib dirname($0);
use BackRest::Archive;
use BackRest::BackupCommon;
use BackRest::BackupFile;
use BackRest::BackupInfo;
use BackRest::Config;
use BackRest::Db;
use BackRest::Exception;
use BackRest::File;
use BackRest::Ini;
use BackRest::Manifest;
use BackRest::ThreadGroup;
use BackRest::Utility;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
our @EXPORT = qw(backup_init backup_cleanup backup backup_expire archive_list_get);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $oDb;
my $oFile;
my $strType; # Type of backup: full, differential (diff), incremental (incr)
my $bCompress;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $bHardLink;
my $bNoStartStop;
my $bForce;
my $iThreadMax;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $iThreadTimeout;
####################################################################################################################################
# BACKUP_INIT
####################################################################################################################################
sub backup_init
{
my $oFileParam = shift;
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
my $oDbParam = shift;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $strTypeParam = shift;
my $bCompressParam = shift;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $bHardLinkParam = shift;
my $iThreadMaxParam = shift;
my $iThreadTimeoutParam = shift;
my $bNoStartStopParam = shift;
my $bForceParam = shift;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$oFile = $oFileParam;
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
$oDb = $oDbParam;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$strType = $strTypeParam;
$bCompress = $bCompressParam;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$bHardLink = $bHardLinkParam;
$iThreadMax = $iThreadMaxParam;
$iThreadTimeout = $iThreadTimeoutParam;
$bNoStartStop = $bNoStartStopParam;
$bForce = $bForceParam;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
####################################################################################################################################
# BACKUP_CLEANUP
####################################################################################################################################
sub backup_cleanup
{
undef($oFile);
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
####################################################################################################################################
# BACKUP_TYPE_FIND - Find the last backup depending on the type
####################################################################################################################################
sub backup_type_find
{
my $strType = shift;
my $strBackupClusterPath = shift;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $strDirectory;
if ($strType eq BACKUP_TYPE_INCR)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
$strDirectory = ($oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(1, 1, 1), 'reverse'))[0];
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
if (!defined($strDirectory) && $strType ne BACKUP_TYPE_FULL)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
$strDirectory = ($oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(1, 0, 0), 'reverse'))[0];
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
return $strDirectory;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
####################################################################################################################################
# BACKUP_FILE_NOT_IN_MANIFEST - Find all files in a backup path that are not in the supplied manifest
####################################################################################################################################
sub backup_file_not_in_manifest
{
my $strPathType = shift;
my $oManifest = shift;
my $oAbortedManifest = shift;
my %oFileHash;
$oFile->manifest($strPathType, undef, \%oFileHash);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my @stryFile;
foreach my $strName (sort(keys(%{$oFileHash{name}})))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
# Ignore certain files that will never be in the manifest
if ($strName eq FILE_MANIFEST ||
$strName eq '.')
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
next;
}
if ($strName eq MANIFEST_KEY_BASE)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
if ($oManifest->test(MANIFEST_SECTION_BACKUP_PATH, $strName))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
next;
}
}
elsif ($strName eq MANIFEST_TABLESPACE)
{
my $bFound = false;
foreach my $strPath ($oManifest->keys(MANIFEST_SECTION_BACKUP_PATH))
{
if ($strPath =~ /^$strName\//)
{
$bFound = true;
last;
}
}
next if $bFound;
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
else
{
my $strBasePath = (split('/', $strName))[0];
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $strPath = substr($strName, length($strBasePath) + 1);
# Create the section from the base path
my $strSection = $strBasePath;
if ($strSection eq 'tablespace')
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
my $strTablespace = (split('/', $strPath))[0];
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$strSection = $strSection . '/' . $strTablespace;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if ($strTablespace eq $strPath)
{
if ($oManifest->test("${strSection}:path"))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
next;
}
}
$strPath = substr($strPath, length($strTablespace) + 1);
}
my $cType = $oFileHash{name}{"${strName}"}{type};
if ($cType eq 'd')
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
if ($oManifest->test("${strSection}:path", "${strPath}"))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
next;
}
}
elsif ($cType eq 'f')
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
if ($oManifest->test("${strSection}:file", "${strPath}") &&
!$oManifest->test("${strSection}:file", $strPath, MANIFEST_SUBKEY_REFERENCE))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
my $strChecksum = $oAbortedManifest->get("${strSection}:file", $strPath, MANIFEST_SUBKEY_CHECKSUM, false);
if (defined($strChecksum) &&
$oManifest->getNumeric("${strSection}:file", $strPath, MANIFEST_SUBKEY_SIZE) ==
$oFileHash{name}{$strName}{size} &&
$oManifest->getNumeric("${strSection}:file", $strPath, MANIFEST_SUBKEY_TIMESTAMP) ==
$oFileHash{name}{$strName}{modification_time})
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
$oManifest->set("${strSection}:file", $strPath, MANIFEST_SUBKEY_CHECKSUM, $strChecksum);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
next;
}
}
}
}
push @stryFile, $strName;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
return @stryFile;
}
####################################################################################################################################
# BACKUP_TMP_CLEAN
#
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Cleans the temp directory from a previous failed backup so it can be reused
####################################################################################################################################
sub backup_tmp_clean
{
my $oManifest = shift;
my $oAbortedManifest = shift;
&log(INFO, 'cleaning backup tmp path');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Remove the pg_xlog directory since it contains nothing useful for the new backup
if (-e $oFile->path_get(PATH_BACKUP_TMP, 'base/pg_xlog'))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
remove_tree($oFile->path_get(PATH_BACKUP_TMP, 'base/pg_xlog')) or confess &log(ERROR, 'unable to delete tmp pg_xlog path');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
# Remove the pg_tblspc directory since it is trivial to rebuild, but hard to compare
if (-e $oFile->path_get(PATH_BACKUP_TMP, 'base/pg_tblspc'))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
remove_tree($oFile->path_get(PATH_BACKUP_TMP, 'base/pg_tblspc')) or confess &log(ERROR, 'unable to delete tmp pg_tblspc path');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
# Get the list of files that should be deleted from temp
my @stryFile = backup_file_not_in_manifest(PATH_BACKUP_TMP, $oManifest, $oAbortedManifest);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
foreach my $strFile (sort {$b cmp $a} @stryFile)
{
my $strDelete = $oFile->path_get(PATH_BACKUP_TMP, $strFile);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# If a path then delete it, all the files should have already been deleted since we are going in reverse order
if (-d $strDelete)
{
&log(DEBUG, "remove path ${strDelete}");
rmdir($strDelete) or confess &log(ERROR, "unable to delete path ${strDelete}, is it empty?");
}
# Else delete a file
else
{
&log(DEBUG, "remove file ${strDelete}");
unlink($strDelete) or confess &log(ERROR, "unable to delete file ${strDelete}");
}
}
}
####################################################################################################################################
# BACKUP_FILE - Performs the file level backup
#
# Uses the information in the manifest to determine which files need to be copied. Directories and tablespace links are only
# created when needed, except in the case of a full backup or if hardlinks are requested.
####################################################################################################################################
sub backup_file
{
my $strDbClusterPath = shift; # Database base data path
my $oBackupManifest = shift; # Manifest for the current backup
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Variables used for parallel copy
my %oFileCopyMap;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $lFileTotal = 0;
my $lSizeTotal = 0;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Determine whether all paths and links will be created
my $bFullCreate = $bHardLink || $strType eq BACKUP_TYPE_FULL;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Iterate through the path sections of the manifest to backup
foreach my $strPathKey ($oBackupManifest->keys(MANIFEST_SECTION_BACKUP_PATH))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
# Determine the source and destination backup paths
my $strBackupSourcePath; # Absolute path to the database base directory or tablespace to backup
my $strBackupDestinationPath; # Relative path to the backup directory where the data will be stored
$strBackupSourcePath = $oBackupManifest->get(MANIFEST_SECTION_BACKUP_PATH, $strPathKey, MANIFEST_SUBKEY_PATH);
$strBackupDestinationPath = $strPathKey;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Create links for tablespaces
if ($oBackupManifest->test(MANIFEST_SECTION_BACKUP_PATH, $strPathKey, MANIFEST_SUBKEY_LINK) && $bFullCreate)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
$oFile->link_create(PATH_BACKUP_TMP, $strBackupDestinationPath,
PATH_BACKUP_TMP,
'base/pg_tblspc/' . $oBackupManifest->get(MANIFEST_SECTION_BACKUP_PATH,
$strPathKey, MANIFEST_SUBKEY_LINK),
false, true, true);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
# If this is a full backup or hard-linked then create all paths and links
if ($bFullCreate)
{
# Create paths
my $strSectionPath = "$strPathKey:path";
if ($oBackupManifest->test($strSectionPath))
{
foreach my $strPath ($oBackupManifest->keys($strSectionPath))
{
if ($strPath ne '.')
{
$oFile->path_create(PATH_BACKUP_TMP, "${strBackupDestinationPath}/${strPath}");
}
}
}
# Create links
my $strSectionLink = "$strPathKey:link";
if ($oBackupManifest->test($strSectionLink))
{
foreach my $strLink ($oBackupManifest->keys($strSectionLink))
{
# Create links except in pg_tblspc because they have already been created
if (!($strPathKey eq 'base' && $strLink =~ /^pg_tblspc\/.*/))
{
$oFile->link_create(PATH_BACKUP_ABSOLUTE,
$oBackupManifest->get($strSectionLink, $strLink, MANIFEST_SUBKEY_DESTINATION),
PATH_BACKUP_TMP, "${strBackupDestinationPath}/${strLink}",
false, false, false);
}
}
}
}
# Possible for the file section to exist with no files (i.e. empty tablespace)
my $strSectionFile = "$strPathKey:file";
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Iterate through the files for each backup source path
foreach my $strFile ($oBackupManifest->keys($strSectionFile))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
my $strBackupSourceFile = "${strBackupSourcePath}/${strFile}";
# If the file has a reference it does not need to be copied since it can be retrieved from the referenced backup.
# However, if hard-linking is turned on the link will need to be created
my $bProcess = true;
my $strReference = $oBackupManifest->get($strSectionFile, $strFile, MANIFEST_SUBKEY_REFERENCE, false);
if (defined($strReference))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
# If hardlinking is turned on then create a hardlink for files that have not changed since the last backup
if ($bHardLink)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
&log(DEBUG, "hardlink ${strBackupSourceFile} to ${strReference}");
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$oFile->link_create(PATH_BACKUP_CLUSTER, "${strReference}/${strBackupDestinationPath}/${strFile}",
PATH_BACKUP_TMP, "${strBackupDestinationPath}/${strFile}", true, false, true);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
else
{
&log(DEBUG, "reference ${strBackupSourceFile} to ${strReference}");
}
$bProcess = false;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
if ($bProcess)
{
my $lFileSize = $oBackupManifest->getNumeric($strSectionFile, $strFile, MANIFEST_SUBKEY_SIZE);
# Setup variables needed for threaded copy
$lFileTotal++;
$lSizeTotal += $lFileSize;
my $strFileKey = sprintf("%016d-${strFile}", $lFileSize);
$oFileCopyMap{$strPathKey}{$strFileKey}{db_file} = $strBackupSourceFile;
$oFileCopyMap{$strPathKey}{$strFileKey}{file_section} = $strSectionFile;
$oFileCopyMap{$strPathKey}{$strFileKey}{file} = ${strFile};
$oFileCopyMap{$strPathKey}{$strFileKey}{backup_file} = "${strBackupDestinationPath}/${strFile}";
$oFileCopyMap{$strPathKey}{$strFileKey}{size} = $lFileSize;
$oFileCopyMap{$strPathKey}{$strFileKey}{modification_time} =
$oBackupManifest->getNumeric($strSectionFile, $strFile, MANIFEST_SUBKEY_TIMESTAMP, false);
$oFileCopyMap{$strPathKey}{$strFileKey}{checksum} =
$oBackupManifest->get($strSectionFile, $strFile, MANIFEST_SUBKEY_CHECKSUM, false);
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
}
# If there are no files to backup then we'll exit with a warning unless in test mode. The other way this could happen is if
# the database is down and backup is called with --no-start-stop twice in a row.
if ($lFileTotal == 0)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
if (!optionGet(OPTION_TEST))
{
confess &log(ERROR, "no files have changed since the last backup - this seems unlikely");
}
return;
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Create backup and result queues
my $oResultQueue = Thread::Queue->new();
my @oyBackupQueue;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Variables used for local copy
my $lSizeCurrent = 0; # Running total of bytes copied
my $bCopied; # Was the file copied?
my $lCopySize; # Size reported by copy
my $strCopyChecksum; # Checksum reported by copy
# Determine how often the manifest will be saved
my $lManifestSaveCurrent = 0;
my $lManifestSaveSize = int($lSizeTotal / 100);
if ($lManifestSaveSize < optionGet(OPTION_MANIFEST_SAVE_THRESHOLD))
{
$lManifestSaveSize = optionGet(OPTION_MANIFEST_SAVE_THRESHOLD);
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Iterate all backup files
foreach my $strPathKey (sort (keys %oFileCopyMap))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
if ($iThreadMax > 1)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
$oyBackupQueue[@oyBackupQueue] = Thread::Queue->new();
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
foreach my $strFileKey (sort {$b cmp $a} (keys(%{$oFileCopyMap{$strPathKey}})))
{
my $oFileCopy = $oFileCopyMap{$strPathKey}{$strFileKey};
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if ($iThreadMax > 1)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
$oyBackupQueue[@oyBackupQueue - 1]->enqueue($oFileCopy);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
else
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
# Backup the file
($bCopied, $lSizeCurrent, $lCopySize, $strCopyChecksum) =
backupFile($oFile, $$oFileCopy{db_file}, $$oFileCopy{backup_file}, $bCompress,
$$oFileCopy{checksum}, $$oFileCopy{modification_time},
$$oFileCopy{size}, $lSizeTotal, $lSizeCurrent);
$lManifestSaveCurrent = backupManifestUpdate($oBackupManifest, $$oFileCopy{file_section}, $$oFileCopy{file},
$bCopied, $lCopySize, $strCopyChecksum, $lManifestSaveSize,
$lManifestSaveCurrent);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
}
}
# If multi-threaded then create threads to copy files
if ($iThreadMax > 1)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
for (my $iThreadIdx = 0; $iThreadIdx < $iThreadMax; $iThreadIdx++)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
my %oParam;
$oParam{compress} = $bCompress;
$oParam{size_total} = $lSizeTotal;
$oParam{queue} = \@oyBackupQueue;
$oParam{result_queue} = $oResultQueue;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
threadGroupRun($iThreadIdx, 'backup', \%oParam);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
# Complete thread queues
my $bDone = false;
do
{
$bDone = threadGroupComplete();
# Read the messages that are passed back from the backup threads
while (my $oMessage = $oResultQueue->dequeue_nb())
{
&log(TRACE, "message received in master queue: section = $$oMessage{file_section}, file = $$oMessage{file}" .
", copied = $$oMessage{copied}");
$lManifestSaveCurrent = backupManifestUpdate($oBackupManifest, $$oMessage{file_section}, $$oMessage{file},
$$oMessage{copied}, $$oMessage{size}, $$oMessage{checksum},
$lManifestSaveSize, $lManifestSaveCurrent);
}
}
while (!$bDone);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
&log(INFO, 'total backup size: ' . file_size_format($lSizeTotal));
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
####################################################################################################################################
# BACKUP
#
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Performs the entire database backup.
####################################################################################################################################
sub backup
{
my $strDbClusterPath = shift;
my $bStartFast = shift;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Record timestamp start
my $lTimestampStart = time();
# Backup start
&log(INFO, "backup start: type = ${strType}");
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Not supporting remote backup hosts yet
if ($oFile->is_remote(PATH_BACKUP))
{
confess &log(ERROR, 'remote backup host not currently supported');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
if (!defined($strDbClusterPath))
{
confess &log(ERROR, 'cluster data path is not defined');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
&log(DEBUG, "cluster path is $strDbClusterPath");
# Create the cluster backup path
$oFile->path_create(PATH_BACKUP_CLUSTER, undef, undef, true);
# Load or build backup.info
my $oBackupInfo = new BackRest::BackupInfo($oFile->path_get(PATH_BACKUP_CLUSTER));
# Build backup tmp and config
my $strBackupTmpPath = $oFile->path_get(PATH_BACKUP_TMP);
my $strBackupConfFile = $oFile->path_get(PATH_BACKUP_TMP, 'backup.manifest');
# Declare the backup manifest
my $oBackupManifest = new BackRest::Manifest($strBackupConfFile, false);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Find the previous backup based on the type
my $oLastManifest = undef;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $strBackupLastPath = backup_type_find($strType, $oFile->path_get(PATH_BACKUP_CLUSTER));
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if (defined($strBackupLastPath))
{
$oLastManifest = new BackRest::Manifest($oFile->path_get(PATH_BACKUP_CLUSTER) . "/${strBackupLastPath}/backup.manifest");
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
&log(INFO, 'last backup label = ' . $oLastManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_LABEL) .
', version = ' . $oLastManifest->get(INI_SECTION_BACKREST, INI_KEY_VERSION));
}
else
{
if ($strType eq BACKUP_TYPE_DIFF)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
&log(WARN, 'No full backup exists, differential backup has been changed to full');
}
elsif ($strType eq BACKUP_TYPE_INCR)
{
&log(WARN, 'No prior backup exists, incremental backup has been changed to full');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
$strType = BACKUP_TYPE_FULL;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
# Backup settings
$oBackupManifest->set(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_TYPE, undef, $strType);
$oBackupManifest->setNumeric(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_TIMESTAMP_START, undef, $lTimestampStart);
$oBackupManifest->setBool(MANIFEST_SECTION_BACKUP_OPTION, MANIFEST_KEY_COMPRESS, undef, $bCompress);
$oBackupManifest->setBool(MANIFEST_SECTION_BACKUP_OPTION, MANIFEST_KEY_HARDLINK, undef, $bHardLink);
$oBackupManifest->setBool(MANIFEST_SECTION_BACKUP_OPTION, MANIFEST_KEY_START_STOP, undef, !$bNoStartStop);
$oBackupManifest->setBool(MANIFEST_SECTION_BACKUP_OPTION, MANIFEST_KEY_ARCHIVE_COPY, undef,
$bNoStartStop || optionGet(OPTION_BACKUP_ARCHIVE_COPY));
$oBackupManifest->setBool(MANIFEST_SECTION_BACKUP_OPTION, MANIFEST_KEY_ARCHIVE_CHECK, undef,
$bNoStartStop || optionGet(OPTION_BACKUP_ARCHIVE_CHECK));
# Database info
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
my ($fDbVersion, $iControlVersion, $iCatalogVersion, $ullDbSysId) = $oDb->info($oFile, $strDbClusterPath);
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
$oBackupManifest->set(MANIFEST_SECTION_BACKUP_DB, MANIFEST_KEY_DB_VERSION, undef, $fDbVersion);
$oBackupManifest->setNumeric(MANIFEST_SECTION_BACKUP_DB, MANIFEST_KEY_CONTROL, undef, $iControlVersion);
$oBackupManifest->setNumeric(MANIFEST_SECTION_BACKUP_DB, MANIFEST_KEY_CATALOG, undef, $iCatalogVersion);
$oBackupManifest->setNumeric(MANIFEST_SECTION_BACKUP_DB, MANIFEST_KEY_SYSTEM_ID, undef, $ullDbSysId);
$oBackupInfo->check($oBackupManifest);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Start backup (unless no-start-stop is set)
my $strArchiveStart;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if ($bNoStartStop)
{
if ($oFile->exists(PATH_DB_ABSOLUTE, $strDbClusterPath . '/' . FILE_POSTMASTER_PID))
{
if ($bForce)
{
&log(WARN, '--no-start-stop passed and ' . FILE_POSTMASTER_PID . ' exists but --force was passed so backup will ' .
'continue though it looks like the postmaster is running and the backup will probably not be ' .
'consistent');
}
else
{
&log(ERROR, '--no-start-stop passed but ' . FILE_POSTMASTER_PID . ' exists - looks like the postmaster is ' .
'running. Shutdown the postmaster and try again, or use --force.');
exit 1;
}
}
}
else
{
my $strTimestampDbStart;
($strArchiveStart, $strTimestampDbStart) =
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
$oDb->backupStart($oFile, $strDbClusterPath, BACKREST_EXE . ' backup started ' .
timestamp_string_get(undef, $lTimestampStart), $bStartFast);
$oBackupManifest->set(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_ARCHIVE_START, undef, $strArchiveStart);
&log(INFO, "archive start: ${strArchiveStart}");
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Build the backup manifest
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
my $oTablespaceMap = $bNoStartStop ? undef : $oDb->tablespaceMapGet();
$oBackupManifest->build($oFile, $strDbClusterPath, $oLastManifest, $bNoStartStop, $oTablespaceMap);
&log(TEST, TEST_MANIFEST_BUILD);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Check if an aborted backup exists for this stanza
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if (-e $strBackupTmpPath)
{
my $bUsable = false;
my $strType = $oBackupManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_TYPE);
my $strPrior = $oBackupManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_PRIOR, undef, false, '<undef>');
my $strVersion = $oBackupManifest->get(INI_SECTION_BACKREST, INI_KEY_VERSION);
my $strAbortedType = '<undef>';
my $strAbortedPrior = '<undef>';
my $strAbortedVersion = '<undef>';
my $oAbortedManifest;
# Attempt to read the manifest file in the aborted backup to see if the backup type and prior backup are the same as the
# new backup that is being started. If any error at all occurs then the backup will be considered unusable and a resume
# will not be attempted.
eval
{
# Load the aborted manifest
$oAbortedManifest = new BackRest::Manifest("${strBackupTmpPath}/backup.manifest");
# Default values if they are not set
$strAbortedType = $oAbortedManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_TYPE);
$strAbortedPrior = $oAbortedManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_PRIOR, undef, false, '<undef>');
$strAbortedVersion = $oAbortedManifest->get(INI_SECTION_BACKREST, INI_KEY_VERSION);
# The backup is usable if between the current backup and the aborted backup:
# 1) The version matches
# 2) The type of both is full or the types match and prior matches
if ($strAbortedVersion eq $strVersion)
{
if ($strAbortedType eq BACKUP_TYPE_FULL
&& $oBackupManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_TYPE) eq BACKUP_TYPE_FULL)
{
$bUsable = true;
}
elsif ($strAbortedType eq $oBackupManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_TYPE) &&
$strAbortedPrior eq $oBackupManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_PRIOR))
{
$bUsable = true;
}
}
};
# If the aborted backup is usable then clean it
if ($bUsable && optionGet(OPTION_RESUME))
{
&log(WARN, 'aborted backup of same type exists, will be cleaned to remove invalid files and resumed');
&log(TEST, TEST_BACKUP_RESUME);
# Clean the old backup tmp path
backup_tmp_clean($oBackupManifest, $oAbortedManifest);
}
# Else remove it
else
{
my $strReason = "resume is disabled";
if (optionGet(OPTION_RESUME))
{
if ($strVersion eq $strAbortedVersion)
{
if ($strType ne $strAbortedType)
{
$strReason = "new type '${strType}' does not match aborted type '${strAbortedType}'";
}
else
{
$strReason = "new prior '${strPrior}' does not match aborted prior '${strAbortedPrior}'";
}
}
else
{
$strReason = "new version '${strVersion}' does not match aborted version '${strVersion}'";
}
}
&log(WARN, "aborted backup exists, but cannot be resumed (${strReason}) - will be dropped and recreated");
&log(TEST, TEST_BACKUP_NORESUME);
remove_tree($oFile->path_get(PATH_BACKUP_TMP))
or confess &log(ERROR, "unable to delete tmp path: ${strBackupTmpPath}");
$oFile->path_create(PATH_BACKUP_TMP);
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
# Else create the backup tmp path
else
{
&log(DEBUG, "creating backup path ${strBackupTmpPath}");
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$oFile->path_create(PATH_BACKUP_TMP);
}
# Save the backup manifest
$oBackupManifest->save();
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Perform the backup
backup_file($strDbClusterPath, $oBackupManifest);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Stop backup (unless no-start-stop is set)
my $strArchiveStop;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if (!$bNoStartStop)
{
my $strTimestampDbStop;
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
($strArchiveStop, $strTimestampDbStop) = $oDb->backupStop();
$oBackupManifest->set(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_ARCHIVE_STOP, undef, $strArchiveStop);
&log(INFO, 'archive stop: ' . $strArchiveStop);
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# If archive logs are required to complete the backup, then check them. This is the default, but can be overridden if the
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# archive logs are going to a different server. Be careful here because there is no way to verify that the backup will be
# consistent - at least not here.
if (!optionGet(OPTION_NO_START_STOP) && optionGet(OPTION_BACKUP_ARCHIVE_CHECK))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
# Save the backup manifest a second time - before getting archive logs in case that fails
$oBackupManifest->save();
# Create the modification time for the archive logs
my $lModificationTime = time();
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# After the backup has been stopped, need to make a copy of the archive logs need to make the db consistent
&log(DEBUG, "retrieving archive logs ${strArchiveStart}:${strArchiveStop}");
my $oArchive = new BackRest::Archive();
my $strArchiveId = $oArchive->getCheck($oFile);
v0.80: DBI Support, Stability, and Convenience Features * Fixed an issue that caused the formatted timestamp for both the oldest and newest backups to be reported as the current time by the info command. Only text output was affected -- json output reported the correct epoch values. Reported by Michael Renner. * Fixed protocol issue that was preventing ssh errors (especially on connection) from being logged. * Now using Perl DBI and DBD::Pg for connections to PostgreSQL rather than psql. The cmd-psql and cmd-psql-option settings have been removed and replaced with db-port and db-socket-path. * Add stop-auto option to allow failed backups to automatically be stopped when a new backup starts. * Add db-timeout option to limit the amount of time pgBackRest will wait for pg_start_backup() and pg_stop_backup() to return. * Remove pg_control file at the beginning of the restore and copy it back at the very end. This prevents the possibility that a partial restore can be started by PostgreSQL. * The repository is now created and updated with consistent directory and file modes. By default umask is set to 0000 but this can be disabled with the neutral-umask setting. * Added checks to be sure the db-path setting is consistent with db-port by comparing the data_directory as reported by the cluster against the db-path setting and the version as reported by the cluster against the value read from pg_control. The db-socket-path setting is checked to be sure it is an absolute path. * Experimental support for PostgreSQL 9.5 alpha1. This may break when the control version or WAL magic changes in future versions but will be updated in each pgBackRest release to keep pace. All regression tests pass except for --target-resume tests (this functionality has changed in 9.5) and there is no testing yet for .partial WAL segments. * Major refactoring of the protocol layer to support future development. * Added vagrant test configurations for Ubuntu 14.04 and CentOS 7. * Split most of README.md out into USERGUIDE.md and CHANGELOG.md because it was becoming unwieldy. Changed most references to "database" in the user guide to "database cluster" for clarity.
2015-08-09 11:20:43 -04:00
my @stryArchive = $oArchive->range($strArchiveStart, $strArchiveStop, $fDbVersion < 9.3);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
foreach my $strArchive (@stryArchive)
{
my $strArchiveFile = $oArchive->walFileName($oFile, $strArchiveId, $strArchive, 600);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if (optionGet(OPTION_BACKUP_ARCHIVE_COPY))
{
&log(DEBUG, "archiving: ${strArchive} (${strArchiveFile})");
# Copy the log file from the archive repo to the backup
my $strDestinationFile = "base/pg_xlog/${strArchive}" . ($bCompress ? ".$oFile->{strCompressExtension}" : '');
my $bArchiveCompressed = $strArchiveFile =~ "^.*\.$oFile->{strCompressExtension}\$";
my ($bCopyResult, $strCopyChecksum, $lCopySize) =
$oFile->copy(PATH_BACKUP_ARCHIVE, "${strArchiveId}/${strArchiveFile}",
PATH_BACKUP_TMP, $strDestinationFile,
$bArchiveCompressed, $bCompress,
undef, $lModificationTime, undef, true);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Add the archive file to the manifest so it can be part of the restore and checked in validation
my $strPathSection = 'base:path';
my $strPathLog = 'pg_xlog';
my $strFileSection = 'base:file';
my $strFileLog = "pg_xlog/${strArchive}";
# Compare the checksum against the one already in the archive log name
if ($strArchiveFile !~ "^${strArchive}-${strCopyChecksum}(\\.$oFile->{strCompressExtension}){0,1}\$")
{
confess &log(ERROR, "error copying WAL segment '${strArchiveFile}' to backup - checksum recorded with " .
"file does not match actual checksum of '${strCopyChecksum}'", ERROR_CHECKSUM);
}
# Set manifest values
$oBackupManifest->set($strFileSection, $strFileLog, MANIFEST_SUBKEY_USER,
$oBackupManifest->get($strPathSection, $strPathLog, MANIFEST_SUBKEY_USER));
$oBackupManifest->set($strFileSection, $strFileLog, MANIFEST_SUBKEY_GROUP,
$oBackupManifest->get($strPathSection, $strPathLog, MANIFEST_SUBKEY_GROUP));
$oBackupManifest->set($strFileSection, $strFileLog, MANIFEST_SUBKEY_MODE, '0700');
$oBackupManifest->set($strFileSection, $strFileLog, MANIFEST_SUBKEY_TIMESTAMP, $lModificationTime);
$oBackupManifest->set($strFileSection, $strFileLog, MANIFEST_SUBKEY_SIZE, $lCopySize);
$oBackupManifest->set($strFileSection, $strFileLog, MANIFEST_SUBKEY_CHECKSUM, $strCopyChecksum);
}
}
}
# Create the path for the new backup
my $lTimestampStop = time();
my $strBackupPath;
if ($strType eq BACKUP_TYPE_FULL || !defined($strBackupLastPath))
{
$strBackupPath = timestamp_file_string_get() . 'F';
$strType = BACKUP_TYPE_FULL;
}
else
{
$strBackupPath = substr($strBackupLastPath, 0, 16);
$strBackupPath .= '_' . timestamp_file_string_get(undef, $lTimestampStop);
if ($strType eq BACKUP_TYPE_DIFF)
{
$strBackupPath .= 'D';
}
else
{
$strBackupPath .= 'I';
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
}
# Record timestamp stop in the config
$oBackupManifest->set(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_TIMESTAMP_STOP, undef, $lTimestampStop + 0);
$oBackupManifest->set(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_LABEL, undef, $strBackupPath);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Save the backup manifest final time
$oBackupManifest->save();
&log(INFO, "new backup label: ${strBackupPath}");
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Rename the backup tmp path to complete the backup
&log(DEBUG, "moving ${strBackupTmpPath} to " . $oFile->path_get(PATH_BACKUP_CLUSTER, $strBackupPath));
$oFile->move(PATH_BACKUP_TMP, undef, PATH_BACKUP_CLUSTER, $strBackupPath);
# Create a link to the most recent backup
$oFile->remove(PATH_BACKUP_CLUSTER, "latest");
$oFile->link_create(PATH_BACKUP_CLUSTER, $strBackupPath, PATH_BACKUP_CLUSTER, "latest", undef, true);
# Save backup info
$oBackupInfo->backupAdd($oFile, $oBackupManifest);
# Backup stop
&log(INFO, 'backup stop');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
####################################################################################################################################
# BACKUP_EXPIRE
#
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Removes expired backups and archive logs from the backup directory. Partial backups are not counted for expiration, so if full
# or differential retention is set to 2, there must be three complete backups before the oldest one can be deleted.
#
# iFullRetention - Optional, must be greater than 0 when supplied.
# iDifferentialRetention - Optional, must be greater than 0 when supplied.
# strArchiveRetention - Optional, must be (full,differential/diff,incremental/incr) when supplied
# iArchiveRetention - Required when strArchiveRetention is supplied. Must be greater than 0.
####################################################################################################################################
sub backup_expire
{
my $strBackupClusterPath = shift; # Base path to cluster backup
my $iFullRetention = shift; # Number of full backups to keep
my $iDifferentialRetention = shift; # Number of differential backups to keep
my $strArchiveRetentionType = shift; # Type of backup to base archive retention on
my $iArchiveRetention = shift; # Number of backups worth of archive to keep
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
my $strPath;
my @stryPath;
# Load or build backup.info
my $oBackupInfo = new BackRest::BackupInfo($oFile->path_get(PATH_BACKUP_CLUSTER));
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Find all the expired full backups
if (defined($iFullRetention))
{
# Make sure iFullRetention is valid
if (!looks_like_number($iFullRetention) || $iFullRetention < 1)
{
confess &log(ERROR, 'full_rentention must be a number >= 1');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
my $iIndex = $iFullRetention;
@stryPath = $oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(1, 0, 0), 'reverse');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
while (defined($stryPath[$iIndex]))
{
# Delete all backups that depend on the full backup. Done in reverse order so that remaining backups will still
# be consistent if the process dies
foreach $strPath ($oFile->list(PATH_BACKUP_CLUSTER, undef, '^' . $stryPath[$iIndex] . '.*', 'reverse'))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
system("rm -rf ${strBackupClusterPath}/${strPath}") == 0
or confess &log(ERROR, "unable to delete backup ${strPath}");
$oBackupInfo->backupRemove($strPath);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
&log(INFO, 'remove expired full backup: ' . $stryPath[$iIndex]);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$iIndex++;
}
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Find all the expired differential backups
if (defined($iDifferentialRetention))
{
# Make sure iDifferentialRetention is valid
if (!looks_like_number($iDifferentialRetention) || $iDifferentialRetention < 1)
{
confess &log(ERROR, 'differential_rentention must be a number >= 1');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
@stryPath = $oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(0, 1, 0), 'reverse');
if (defined($stryPath[$iDifferentialRetention - 1]))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
&log(DEBUG, 'differential expiration based on ' . $stryPath[$iDifferentialRetention - 1]);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Get a list of all differential and incremental backups
foreach $strPath ($oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(0, 1, 1), 'reverse'))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
&log(DEBUG, "checking ${strPath} for differential expiration");
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Remove all differential and incremental backups before the oldest valid differential
if ($strPath lt $stryPath[$iDifferentialRetention - 1])
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
system("rm -rf ${strBackupClusterPath}/${strPath}") == 0
or confess &log(ERROR, "unable to delete backup ${strPath}");
$oBackupInfo->backupRemove($strPath);
&log(INFO, "remove expired diff/incr backup ${strPath}");
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
}
}
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# If no archive retention type is set then exit
if (!defined($strArchiveRetentionType))
{
&log(INFO, 'archive rentention type not set - archive logs will not be expired');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
return;
}
# Determine which backup type to use for archive retention (full, differential, incremental)
if ($strArchiveRetentionType eq BACKUP_TYPE_FULL)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
if (!defined($iArchiveRetention))
{
$iArchiveRetention = $iFullRetention;
}
@stryPath = $oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(1, 0, 0), 'reverse');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
elsif ($strArchiveRetentionType eq BACKUP_TYPE_DIFF)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
if (!defined($iArchiveRetention))
{
$iArchiveRetention = $iDifferentialRetention;
}
@stryPath = $oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(1, 1, 0), 'reverse');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
elsif ($strArchiveRetentionType eq BACKUP_TYPE_INCR)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
@stryPath = $oFile->list(PATH_BACKUP_CLUSTER, undef, backupRegExpGet(1, 1, 1), 'reverse');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
else
{
confess &log(ERROR, "unknown archive_retention_type '${strArchiveRetentionType}'");
}
# Make sure that iArchiveRetention is set and valid
if (!defined($iArchiveRetention))
{
confess &log(ERROR, 'archive_rentention must be set if archive_retention_type is set');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
return;
}
if (!looks_like_number($iArchiveRetention) || $iArchiveRetention < 1)
{
confess &log(ERROR, 'archive_rentention must be a number >= 1');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# if no backups were found then preserve current archive logs - too scary to delete them!
my $iBackupTotal = scalar @stryPath;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if ($iBackupTotal == 0)
{
return;
}
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# See if enough backups exist for expiration to start
my $strArchiveRetentionBackup = $stryPath[$iArchiveRetention - 1];
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if (!defined($strArchiveRetentionBackup))
{
if ($strArchiveRetentionType eq BACKUP_TYPE_FULL && scalar @stryPath > 0)
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
&log(INFO, 'fewer than required backups for retention, but since archive_retention_type = full using oldest full backup');
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
$strArchiveRetentionBackup = $stryPath[scalar @stryPath - 1];
}
if (!defined($strArchiveRetentionBackup))
{
return;
}
}
# Get the archive logs that need to be kept. To be cautious we will keep all the archive logs starting from this backup
# even though they are also in the pg_xlog directory (since they have been copied more than once).
&log(INFO, 'archive retention based on backup ' . $strArchiveRetentionBackup);
my $oManifest = new BackRest::Manifest($oFile->path_get(PATH_BACKUP_CLUSTER) . "/${strArchiveRetentionBackup}/backup.manifest");
my $strArchiveLast = $oManifest->get(MANIFEST_SECTION_BACKUP, MANIFEST_KEY_ARCHIVE_START);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
if (!defined($strArchiveLast))
{
confess &log(ERROR, "invalid archive location retrieved ${strArchiveRetentionBackup}");
}
&log(INFO, 'archive retention starts at ' . $strArchiveLast);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Get archive info
my $oArchive = new BackRest::Archive();
my $strArchiveId = $oArchive->getCheck($oFile);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Remove any archive directories or files that are out of date
foreach $strPath ($oFile->list(PATH_BACKUP_ARCHIVE, $strArchiveId, "^[0-F]{16}\$"))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
&log(DEBUG, 'found major archive path ' . $strPath);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# If less than first 16 characters of current archive file, then remove the directory
if ($strPath lt substr($strArchiveLast, 0, 16))
{
my $strFullPath = $oFile->path_get(PATH_BACKUP_ARCHIVE, $strArchiveId) . "/${strPath}";
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
remove_tree($strFullPath) > 0 or confess &log(ERROR, "unable to remove ${strFullPath}");
&log(DEBUG, 'remove major archive path ' . $strFullPath);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
# If equals the first 16 characters of the current archive file, then delete individual files instead
elsif ($strPath eq substr($strArchiveLast, 0, 16))
{
my $strSubPath;
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
# Look for archive files in the archive directory
foreach $strSubPath ($oFile->list(PATH_BACKUP_ARCHIVE, "${strArchiveId}/${strPath}", "^[0-F]{24}.*\$"))
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
{
# Delete if the first 24 characters less than the current archive file
if ($strSubPath lt substr($strArchiveLast, 0, 24))
{
unlink($oFile->path_get(PATH_BACKUP_ARCHIVE, "${strArchiveId}/${strSubPath}"))
or confess &log(ERROR, 'unable to remove ' . $strSubPath);
&log(DEBUG, 'remove expired archive file ' . $strSubPath);
v0.10: Backup and archiving are functional This version has been put into production at Resonate, so it does work, but there are a number of major caveats. * No restore functionality, but the backup directories are consistent Postgres data directories. You'll need to either uncompress the files or turn off compression in the backup. Uncompressed backups on a ZFS (or similar) filesystem are a good option because backups can be restored locally via a snapshot to create logical backups or do spot data recovery. * Archiving is single-threaded. This has not posed an issue on our multi-terabyte databases with heavy write volume. Recommend a large WAL volume or to use the async option with a large volume nearby. * Backups are multi-threaded, but the Net::OpenSSH library does not appear to be 100% threadsafe so it will very occasionally lock up on a thread. There is an overall process timeout that resolves this issue by killing the process. Yes, very ugly. * Checksums are lost on any resumed backup. Only the final backup will record checksum on multiple resumes. Checksums from previous backups are correctly recorded and a full backup will reset everything. * The backup.manifest is being written as Storable because Config::IniFile does not seem to handle large files well. Would definitely like to save these as human-readable text. * Absolutely no documentation (outside the code). Well, excepting these release notes. * Lots of other little things and not so little things. Much refactoring to follow.
2014-03-05 19:53:13 -05:00
}
}
}
}
}
1;