Warning: This document is for the development version of Bareos Documentation. The main version is bareos-18.2.

Howtos

Use a dummy device to test the backup

If your are testing your configuration, but don’t want to store the backup data, it is possible to use a dummy FIFO device to test your configuration, see Stored configuration.

Obviously, it can not be used to do a restore.

FIFO Storage Device Configuration
Device {
  Name = NULL
  Media Type = NULL
  Device Type = Fifo
  Archive Device = /dev/null
  LabelMedia = yes
  Random Access = no
  AutomaticMount = no
  RemovableMedia = no
  MaximumOpenWait = 60
  AlwaysOpen = no
}

Backup Of Third Party Databases

If you are running a database in production mode on your machine, Bareos will happily backup the files, but if the database is in use while Bareos is reading it, you may back it up in an unstable state.

The best solution is to shutdown your database before backing it up, or use some tool specific to your database to make a valid live copy perhaps by dumping the database in ASCII format.

Backup of MSSQL Databases with Bareos Plugin

Version >= 13.2.0

Preparation

If you like to use the MSSQL-Plugin to backing up your Databases you need to consider some things:

  • Database Mode
    The database need to run in Full Recovery Mode. Otherwise you are not able to use differential and incremental backups or to use point in time recovery.

Warning

If you set the databases into the mentionend mode you have to consider some maintance facts. The database doesn’t shrink or delete the logs unanttended, so you have to shrink them manual once a week and you have to truncate the logs once in a month.

  • Security and Access
    For connections you can use a SQL-User or a integrated systemaccount (Windows NT user). Both connection types are supported.
    • Standard Security
      You have to provide user credentials within your options which do belong to user with the sufficent right performing restores and backups from the database. This way stands for an extra backup/restore user.
    • Trusted Security
      You use a systemaccount which have the sufficent rights to performing backups and restores on a database. This systemaccount have to be same account like the bareos-filedeamon runs on.
  • Permissions and Roles

    • Server-Role
      The user should be in the groups sysadmin and dbcreator.
    • Database permissions
      The user have to be a backupoperator and dbowner of the database which you like to backup.

There is no difference for the rights and roles between using a systemaccount (trusted security method) or a extra backup user (standard security method). Please keep in mind if you use the trusted security method you have to use the same system account like the bareos-filedeamon runs on.

MSSQL Plugin Installation

For Bareos < 14.2, install the Bareos MSSQL plugin onto the MSSQL server you want to backup. Bareos >= 14.2 also allows to backup remote MSSQL servers (option serveraddress).

Bareos Windows-Installer

Install the Bareos filedaemon including the component “Bareos FileDameon Plugins”. Make sure, that you install the file daemon without the “compatible” option.

Manual install

After downloading the plugin you need to copy it into C:Program FilesBareosPlugins. Then you need to define the plugin directory and which plugin the bareos-filedaemon should use. You have to edit the bareos-filedaemon resource in C:Program Databareos-fd.conf as follows:

MSSQL plugin configuration
FileDaemon {
  Name = mssqlserver-fd
  Maximum Concurrent Jobs = 20

  # remove comment in next line to load plugins from specified directory
  Plugin Directory = "C:/Program Files/Bareos/Plugins"

  Plugin Names = "mssqlvdi"
  compatible = no  # this is the default since bareos 15
}

Plugin Test

status client=mssqlserver-fd
*status client=mssqlserver-fd
Connecting to Client mssqlserver-fd at 192.168.10.101:9102

mssqlserver-fd Version: 13.2.2 (12 November 2013)  VSS Linux Cross-compile Win64
Daemon started 18-Nov-13 11:51. Jobs: run=0 running=0.
Microsoft Windows Server 2012 Standard Edition (build 9200), 64-bit
 Heap: heap=0 smbytes=20,320 max_bytes=20,522 bufs=71 max_bufs=73
 Sizeof: boffset_t=8 size_t=8 debug=0 trace=1 bwlimit=0kB/s
Plugin Info:
 Plugin     : mssqlvdi-fd.dll
 Description: Bareos MSSQL VDI Windows File Daemon Plugin
 Version    : 1, Date: July 2013
 Author     : Zilvinas Krapavickas
 License    : Bareos AGPLv3
 Usage      :
  mssqlvdi:
  serveraddress=<hostname>:
  instance=<instance name>:
  database=<database name>:
  username=<database username>:
  password=<database password>:
  norecovery=<yes|no>:
  replace=<yes|no>:
  recoverafterrestore=<yes|no>:
  stopbeforemark=<log sequence number specification>:
  stopatmark=<log sequence number specification>:
  stopat=<timestamp>

 examples:
  timestamp: 'Apr 15, 2020 12:00 AM'
  log sequence number: 'lsn:15000000040000037'

Configure the FileSet

To use the plugin you need to configure it in the fileset as a plugin resource. For each database instance you need to define a exclusive backup job and fileset.

MSSQL FileSet
Fileset {
  Name = "Mssql"
  Enable VSS = no
  Include {
    Options {
      Signature = MD5
    }
    Plugin = "mssqlvdi:instance=default:database=myDatabase:username=bareos:password=bareos"
  }
}

In this example we use the standard security method for the connection.

Used options in the plugin string are:

mssqlvdi
This is the reference to the MSSQL plugin.
serveraddress
(Version >= 14.2.2) Defines the server address to connect to (if empty defaults to localhost).
instance
Defines the instance within the database server.
database
Defines the database that should get backed up.
username and password
Username and Password are required, when the connection is done using a MSSQL user. If the systemaccount the bareos-fd runs with has succifient permissions, this is not required.

It is recommend to define an additional restore job.

For every database separate job and FileSet are required.

Run Backups

Here you can see an example for a backup:

run MSSQL backup job
*run job=MSSQLBak
Using Catalog "MyCatalog"
Run Backup job
JobName:  MSSQLBak
Level:    Full
Client:   mssqlserver-fd
Format:   Native
FileSet:  Mssql
Pool:     File (From Job resource)
Storage:  File (From Job resource)
When:     2013-11-21 09:48:27
Priority: 10
OK to run? (yes/mod/no): yes
Job queued. JobId=7
You have no messages.
*mess
21-Nov 09:48 bareos-dir JobId 7: Start Backup JobId 7, Job=MSSQLBak.2013-11-21_09.48.30_04
21-Nov 09:48 bareos-dir JobId 7: Using Device "FileStorage" to write.
21-Nov 09:49 bareos-sd JobId 7: Volume "test1" previously written, moving to end of data.
21-Nov 09:49 bareos-sd JobId 7: Ready to append to end of Volume "test1" size=2300114868
21-Nov 09:49 bareos-sd JobId 7: Elapsed time=00:00:27, Transfer rate=7.364 M Bytes/second

21-Nov 09:49 bareos-dir JobId 7: Bareos bareos-dir 13.4.0 (01Oct13):
  Build OS:               x86_64-pc-linux-gnu debian Debian GNU/Linux 7.0 (wheezy)
  JobId:                  7
  Job:                    MSSQLBak.2013-11-21_09.48.30_04
  Backup Level:           Full
  Client:                 "mssqlserver-fd" 13.2.2 (12Nov13) Microsoft Windows Server 2012 Standard Edition (build 9200), 64-bit,Cross-compile,Win64
  FileSet:                "Mssql" 2013-11-04 23:00:01
  Pool:                   "File" (From Job resource)
  Catalog:                "MyCatalog" (From Client resource)
  Storage:                "File" (From Job resource)
  Scheduled time:         21-Nov-2013 09:48:27
  Start time:             21-Nov-2013 09:49:13
  End time:               21-Nov-2013 09:49:41
  Elapsed time:           28 secs
  Priority:               10
  FD Files Written:       1
  SD Files Written:       1
  FD Bytes Written:       198,836,224 (198.8 MB)
  SD Bytes Written:       198,836,435 (198.8 MB)
  Rate:                   7101.3 KB/s
  Software Compression:   None
  VSS:                    no
  Encryption:             no
  Accurate:               no
  Volume name(s):         test1
  Volume Session Id:      1
  Volume Session Time:    1384961357
  Last Volume Bytes:      2,499,099,145 (2.499 GB)
  Non-fatal FD errors:    0
  SD Errors:              0
  FD termination status:  OK
  SD termination status:  OK
  Termination:            Backup OK

At least you gain a full backup which contains the follow:

@MSSQL/
@MSSQL/default/
@MSSQL/default/myDatabase/
@MSSQL/default/myDatabase/db-full
So if you perform your first full backup your are capable to perfom differntial and incremental backups.
Differntial FileSet example:
/@MSSQL/
/@MSSQL/default/
/@MSSQL/default/myDatabase/
/@MSSQL/default/myDatabase/db-full
/@MSSQL/default/myDatabase/db-diff

Incremental FileSet example:

*@MSSQL/
  *default/
    *myDatabase/
      *db-diff
      *db-full
      *log-2013-11-21 17:32:20

Restores

If you want to perfom a restore of a full backup without differentials or incrementals you have some options which helps you to restore even the corrupted database still exist. But you have to specifiy the options like plugin, instance and database during every backup.

replace=<yes|no>
With this option you can replace the database if it still exist.
instance
Defines the server instance whithin the database is running.
database
Defines the database you want to backup.

If you want to restore the actual backup to a set of backup files which you can use to restore a database under an new name or perform any kind of special operations using for example the sql management studio, you can use a where setting for the restore other then ’/’. When the where is set to ’/’ it will restore to the Virtual Device Interface (VDI).

When you specify for restore a where path which is lets say ’c:/temp’ the plugin will restore the selected backup files under a relocated path under c:/temp/@MSSQL@/…

Example for a full restore:

restore MSSQL database
*restore client=mssqlserver-fd
Using Catalog "MyCatalog"

First you select one or more JobIds that contain files
to be restored. You will be presented several methods
of specifying the JobIds. Then you will be allowed to
select which files from those JobIds are to be restored.

To select the JobIds, you have the following choices:
     1: List last 20 Jobs run
     2: List Jobs where a given File is saved
     3: Enter list of comma separated JobIds to select
     4: Enter SQL list command
     5: Select the most recent backup for a client
     6: Select backup for a client before a specified time
     7: Enter a list of files to restore
     8: Enter a list of files to restore before a specified time
     9: Find the JobIds of the most recent backup for a client
    10: Find the JobIds for a backup for a client before a specified time
    11: Enter a list of directories to restore for found JobIds
    12: Select full restore to a specified Job date
    13: Cancel
Select item:  (1-13): 5
Automatically selected FileSet: Mssql
+-------+-------+----------+-------------+---------------------+------------+
| JobId | Level | JobFiles | JobBytes    | StartTime           | VolumeName |
+-------+-------+----------+-------------+---------------------+------------+
|     8 | F     |        1 | 198,836,224 | 2013-11-21 09:52:28 | test1      |
+-------+-------+----------+-------------+---------------------+------------+
You have selected the following JobId: 8

Building directory tree for JobId(s) 8 ...
1 files inserted into the tree.

You are now entering file selection mode where you add (mark) and
remove (unmark) files to be restored. No files are initially added, unless
you used the "all" keyword on the command line.
Enter "done" to leave this mode.

cwd is: /
$ mark *
1 file marked.
$ done
Bootstrap records written to /var/lib/bareos/bareos-dir.restore.4.bsr

The job will require the following
   Volume(s)                 Storage(s)                SD Device(s)
===========================================================================

    test1                     File                      FileStorage

Volumes marked with "*" are online.


1 file selected to be restored.

The defined Restore Job resources are:
     1: RestoreMSSQL
     2: RestoreFiles
Select Restore Job (1-2): 1
Using Catalog "MyCatalog"
Run Restore job
JobName:         RestoreMSSQL
Bootstrap:       /var/lib/bareos/bareos-dir.restore.4.bsr
Where:           /
Replace:         Always
FileSet:         Mssql
Backup Client:   mssqlserver-fd
Restore Client:  mssqlserver-fd
Format:          Native
Storage:         File
When:            2013-11-21 17:12:05
Catalog:         MyCatalog
Priority:        10
Plugin Options:  *None*
OK to run? (yes/mod/no): mod
Parameters to modify:
     1: Level
     2: Storage
     3: Job
     4: FileSet
     5: Restore Client
     6: Backup Format
     7: When
     8: Priority
     9: Bootstrap
    10: Where
    11: File Relocation
    12: Replace
    13: JobId
    14: Plugin Options
Select parameter to modify (1-14): 14
Please enter Plugin Options string: mssqlvdi:instance=default:database=myDatabase:replace=yes
Run Restore job
JobName:         RestoreMSSQL
Bootstrap:       /var/lib/bareos/bareos-dir.restore.4.bsr
Where:           /
Replace:         Always
FileSet:         Mssql
Backup Client:   mssqlserver-fd
Restore Client:  mssqlserver-fd
Format:          Native
Storage:         File
When:            2013-11-21 17:12:05
Catalog:         MyCatalog
Priority:        10
Plugin Options:  mssqlvdi:instance=default:database=myDatabase:replace=yes
OK to run? (yes/mod/no): yes
Job queued. JobId=10
You have messages.
*mess
21-Nov 17:12 bareos-dir JobId 10: Start Restore Job RestoreMSSQL.2013-11-21_17.12.26_11
21-Nov 17:12 bareos-dir JobId 10: Using Device "FileStorage" to read.
21-Nov 17:13 damorgan-sd JobId 10: Ready to read from volume "test1" on device "FileStorage" (/storage).
21-Nov 17:13 damorgan-sd JobId 10: Forward spacing Volume "test1" to file:block 0:2499099145.
21-Nov 17:13 damorgan-sd JobId 10: End of Volume at file 0 on device "FileStorage" (/storage), Volume "test1"
21-Nov 17:13 damorgan-sd JobId 10: End of all volumes.
21-Nov 17:13 bareos-dir JobId 10: Bareos bareos-dir 13.4.0 (01Oct13):
  Build OS:               x86_64-pc-linux-gnu debian Debian GNU/Linux 7.0 (wheezy)
  JobId:                  10
  Job:                    RestoreMSSQL.2013-11-21_17.12.26_11
  Restore Client:         mssqlserver-fd
  Start time:             21-Nov-2013 17:12:28
  End time:               21-Nov-2013 17:13:21
  Files Expected:         1
  Files Restored:         1
  Bytes Restored:         198,836,224
  Rate:                   3751.6 KB/s
  FD Errors:              0
  FD termination status:  OK
  SD termination status:  OK
  Termination:            Restore OK
Restore a Backup Chain

If you like to restore a specific state or a whole chain consists of full, incremental and differential backups you need to use the “norecovery=yes” option. After this the database is in “recovery mode”. You can also use a option which put the database right after the restore back into the right mode. If you like to restore certains point with protocols or “LSN” it it not recommend to work with this option.

norecovery=<yes|no>
This option must be set to yes, if the database server should not do a automatic recovery after the backup. Instead, additional manual maintenace operations are possible.
recoverafterrestore=<yes|no>

With this command the database is right after backup in the correct mode. If you not use this you have to use the followed tsql statement:

Restore DATABASE yourDatabase WITH RECOVERY
GO
stopbeforemark=<log sequence number specification>
used for point in time recovery.
stopatmark=<log sequence number specification>
used for point in time recovery.
stopat=<timestamp>
used for point in time recovery.

Followed is a example for a restore of full, differential and incremental backup with a replace of the original database:

restore MSSQL database chain
*restore client=mssqlserver-fd

First you select one or more JobIds that contain files
to be restored. You will be presented several methods
of specifying the JobIds. Then you will be allowed to
select which files from those JobIds are to be restored.

To select the JobIds, you have the following choices:
     1: List last 20 Jobs run
     2: List Jobs where a given File is saved
     3: Enter list of comma separated JobIds to select
     4: Enter SQL list command
     5: Select the most recent backup for a client
     6: Select backup for a client before a specified time
     7: Enter a list of files to restore
     8: Enter a list of files to restore before a specified time
     9: Find the JobIds of the most recent backup for a client
    10: Find the JobIds for a backup for a client before a specified time
    11: Enter a list of directories to restore for found JobIds
    12: Select full restore to a specified Job date
    13: Cancel
Select item:  (1-13): 5
Automatically selected FileSet: Mssql
+-------+-------+----------+-------------+---------------------+------------+
| JobId | Level | JobFiles | JobBytes    | StartTime           | VolumeName |
+-------+-------+----------+-------------+---------------------+------------+
|     8 | F     |        1 | 198,836,224 | 2013-11-21 09:52:28 | test1      |
|    11 | D     |        1 |   2,555,904 | 2013-11-21 17:19:45 | test1      |
|    12 | I     |        1 |     720,896 | 2013-11-21 17:29:39 | test1      |
+-------+-------+----------+-------------+---------------------+------------+
You have selected the following JobIds: 8,11,12

Building directory tree for JobId(s) 8,11,12 ...
3 files inserted into the tree.

You are now entering file selection mode where you add (mark) and
remove (unmark) files to be restored. No files are initially added, unless
you used the "all" keyword on the command line.
Enter "done" to leave this mode.

cwd is: /
$ mark *
3 files marked.
$ lsmark
*@MSSQL/
  *default/
    *myDatabase/
      *db-diff
      *db-full
      *log-2013-11-21 17:32:20
$ done
Bootstrap records written to /var/lib/bareos/bareos-dir.restore.6.bsr

The job will require the following
   Volume(s)                 Storage(s)                SD Device(s)
===========================================================================

    test1                     File                      FileStorage

Volumes marked with "*" are online.


1 file selected to be restored.

The defined Restore Job resources are:
     1: RestoreMSSQL
     2: RestoreFiles
Select Restore Job (1-2): 1
Run Restore job
JobName:         RestoreMSSQL
Bootstrap:       /var/lib/bareos/bareos-dir.restore.6.bsr
Where:           /
Replace:         Always
FileSet:         Mssql
Backup Client:   mssqlserver-fd
Restore Client:  mssqlserver-fd
Format:          Native
Storage:         File
When:            2013-11-21 17:34:23
Catalog:         MyCatalog
Priority:        10
Plugin Options:  *None*
OK to run? (yes/mod/no): mod
Parameters to modify:
     1: Level
     2: Storage
     3: Job
     4: FileSet
     5: Restore Client
     6: Backup Format
     7: When
     8: Priority
     9: Bootstrap
    10: Where
    11: File Relocation
    12: Replace
    13: JobId
    14: Plugin Options
Select parameter to modify (1-14): 14
Please enter Plugin Options string: mssqlvdi:instance=default:database=myDatabase:replace=yes:norecovery=yes
Run Restore job
JobName:         RestoreMSSQL
Bootstrap:       /var/lib/bareos/bareos-dir.restore.6.bsr
Where:           /
Replace:         Always
FileSet:         Mssql
Backup Client:   mssqlserver-fd
Restore Client:  mssqlserver-fd
Format:          Native
Storage:         File
When:            2013-11-21 17:34:23
Catalog:         MyCatalog
Priority:        10
Plugin Options:  mssqlvdi:instance=default:database=myDatabase:replace=yes:norecovery=yes
OK to run? (yes/mod/no): yes
Job queued. JobId=14
21-Nov 17:34 bareos-dir JobId 14: Start Restore Job RestoreMSSQL.2013-11-21_17.34.40_16
21-Nov 17:34 bareos-dir JobId 14: Using Device "FileStorage" to read.
21-Nov 17:35 damorgan-sd JobId 14: Ready to read from volume "test1" on device "FileStorage" (/storage).
21-Nov 17:35 damorgan-sd JobId 14: Forward spacing Volume "test1" to file:block 0:2499099145.
21-Nov 17:35 damorgan-sd JobId 14: End of Volume at file 0 on device "FileStorage" (/storage), Volume "test1"
21-Nov 17:35 damorgan-sd JobId 14: End of all volumes.
21-Nov 17:35 bareos-dir JobId 14: Bareos bareos-dir 13.4.0 (01Oct13):
  Build OS:               x86_64-pc-linux-gnu debian Debian GNU/Linux 7.0 (wheezy)
  JobId:                  14
  Job:                    RestoreMSSQL.2013-11-21_17.34.40_16
  Restore Client:         mssqlserver-fd
  Start time:             21-Nov-2013 17:34:42
  End time:               21-Nov-2013 17:35:36
  Files Expected:         1
  Files Restored:         3
  Bytes Restored:         202,113,024
  Rate:                   3742.8 KB/s
  FD Errors:              0
  FD termination status:  OK
  SD termination status:  OK
  Termination:            Restore OK

Backup of a PostgreSQL Database

In this section, we describe different methods how to do backups of the PostgreSQL databases.

Backup of a PostgreSQL Database by using the RunScript directive

One method to backup a PostgreSQL database is to use the pg_dumpall tool to dump the database into a file and then backup it as a normal file. After the backup, the file can be removed. It may also be an option not to remove it, so that the latest version is always available immediately. On the next job run it will be overwritten anyway.

This can be done by using Run Script (Dir->Job) directives inside a Job Resource, for example:

RunScript job resource for a PostgreSQL backup
Job {
  Name = "BackupDatabase"
  JobDefs = "DefaultJob"
  Client = dbserver-fd
  Level = Full
  FileSet="Database"

  # This creates a dump of our database in the local filesystem on the client
  RunScript {
    FailJobOnError = Yes
    RunsOnClient = Yes
    RunsWhen = Before
    Command = "sh -c 'pg_dumpall -U postgres > /var/lib/bareos/postgresql_dump.sql'"
  }

  # This deletes the dump in our local filesystem on the client
  RunScript {
    RunsOnSuccess = Yes
    RunsOnClient = Yes
    RunsWhen = After
    Command = "rm /var/lib/bareos/postgresql_dump.sql"
  }
}

FileSet {
  Name = "Database"
  Include {
    Options {
      signature = MD5
      compression = gzip
    }
    # database dump file
    File = "/var/lib/bareos/postgresql_dump.sql"
  }
}

Note that redirecting the pg_dumpall output to a file requires to run the whole command line through a shell, otherwise the pg_dumpall would not know what do with the > character and the job would fail. As no shell features like redirection or piping are used for the rm, the sh -c is not needed there. See Run Script (Dir->Job) for more details.

Backup of a PostgreSQL Databases by using the bpipe plugin

Instead of creating a temporary database dump file, the bpipe plugin can be used. For general information about bpipe, see the bpipe Plugin section. The bpipe plugin is configured inside the Include (Dir->Fileset) section of a File Set, e.g.:

bpipe directive for PostgreSQL backup
FileSet {
  Name = "postgresql-all"
  Include {
    Plugin = "bpipe:file=/POSTGRESQL/dump.sql:reader=pg_dumpall -U postgres:writer=psql -U postgres"
    Options {
      signature = MD5
      compression = gzip
    }
  }
}

This causes the File Daemon to call bpipe plugin, which will write its data into the “pseudo” file /POSTGRESQL/dump.sql by calling the program pg_dumpall -U postgres to read the data during backup. The pg_dumpall command outputs all the data for the database, which will be read by the plugin and stored in the backup. During restore, the data that was backed up will be sent to the program specified in the last field, which in this case is psql. When psql is called, it will read the data sent to it by the plugin then write it back to the same database from which it came from.

This can also be used, to backup a database that is running on a remote host:

bpipe directive to backup a PostgreSQL database that is running on a remote host
FileSet {
  Name = "postgresql-remote"
  Include {
    Plugin = "bpipe:file=/POSTGRESQL/dump.sql:reader=pg_dumpall -h <hostname> -U <username> -W <password>:writer=psql -h <hostname> -U <username> -W <password>"
    Options {
      signature = MD5
      compression = gzip
    }
  }
}

Backup of a PostgreSQL Databases by using the PGSQL-Plugin

The PGSQL-Plugin supports an online (Hot) backup of database files and database transaction logs (WAL) archiving (with pgsql-archlog) and backup. With online database and transaction logs the backup plugin can perform Poin-In-Time-Restore up to a single selected transaction or date/time.

Database recovery is performed fully automatic with dedicated pgsql-restore utility.

For a full description, see https://github.com/bareos/contrib-pgsql-plugin/wiki.

Backup of a MySQL Database

In this section, we describe different methods to do a full backup of a MySQL database.

Backup of MySQL Databases using the Bareos MySQL Percona xtrabackup Plugin

This plugin is available since Version >= 16.2.4, it uses the xtrabackup tool from Percona to perform full and incremental hot-backups of MySQL / MariaDB tables of type InnoDB. It can also backup MyISAM tables but only as full backups. On restore it requires a preparation using the xtrabackup tools, before the tables can be restored. If you simply want to backup full dumps, then using Backup of MySQL Databases using the Python MySQL plugin is the easier way.

Prerequisites

Install the xtrabackup tool from Percona. Documentation and packages are available here: https://www.percona.com/software/mysql-database/percona-xtrabackup. The plugin was successfully tested with xtrabackup versions 2.3.5 and 2.4.4.

As it is a Python plugin, it will also require to have the package bareos-filedaemon-python-plugin installed on the Bareos File Daemon, where you run it.

For authentication the .mycnf file of the user running the Bareos File Daemon. Before proceeding, make sure that xtrabackup can connect to the database and create backups.

Installation

Make sure you have met the prerequisites. Install the files BareosFdPercona.py and bareos-fd-percona.py in your Bareos plugin directory (usually /usr/lib64/bareos/plugins). These files are available in the Git repository https://github.com/bareos/bareos-contrib/tree/master/fd-plugins/bareos_percona.

Configuration

Activate your plugin directory in the Bareos File Daemon configuration. See File Daemon Plugins for more about plugins in general.

bareos-fd.d/client/myself.conf
Client {
  ...
  Plugin Directory = /usr/lib64/bareos/plugins
  Plugin Names = "python"
}

Now include the plugin as command-plugin in the Fileset resource:

bareos-dir.d/fileset/mysql.conf
FileSet {
    Name = "mysql"
    Include  {
        Options {
            compression=GZIP
            signature = MD5
        }
        File = /etc
        #...
        Plugin = "python:module_path=/usr/lib64/bareos/plugins:module_name=bareos-fd-percona:mycnf=/root/.my.cnf"
    }
}

If used this way, the plugin will call xtrabackup to create a backup of all databases in the xbstream format. This stream will be processed by Bareos. If job level is incremental, xtrabackup will perform an incremental backup since the last backup – for InnoDB tables. If you have MyISAM tables, you will get a full backup of those.

You can append options to the plugin call as key=value pairs, separated by ’:’. The following options are available:

  • With mycnf you can make xtrabackup use a special mycnf-file with login credentials.
  • dumpbinary lets you modify the default command xtrabackup.
  • dumpoptions to modify the options for xtrabackup. Default setting is: --backup --datadir=/var/lib/mysql/ --stream=xbstream --extra-lsndir=/tmp/individual_tempdir
  • restorecommand to modify the command for restore. Default setting is: xbstream -x -C
  • strictIncremental: By default (false), an incremental backup will create data, even if the Log Sequence Number (LSN) wasn’t increased since last backup. This is to ensure, that eventual changes to MYISAM tables get into the backup. MYISAM does not support incremental backups, you will always get a full bakcup of these tables. If set to true, no data will be written into backup, if the LSN wasn’t changed.
Restore

With the usual Bareos restore mechanism a file-hierarchy will be created on the restore client under the default restore location:

/tmp/bareos-restores/_percona/

Each restore job gets an own subdirectory, because Percona expects an empty directory. In that subdirectory, a new directory is created for every backup job that was part of the Full-Incremental sequence.

The naming scheme is: fromLSN_toLSN_jobid

Example:

/tmp/bareos-restores/_percona/351/
|-- 00000000000000000000_00000000000010129154_0000000334
|-- 00000000000010129154_00000000000010142295_0000000335
|-- 00000000000010142295_00000000000010201260_0000000338

This example shows the restore tree for restore job with ID 351. First subdirectory has all files from the first full backup job with ID 334. It starts at LSN 0 and goes until LSN 10129154.

Next line is the first incremental job with ID 335, starting at LSN 10129154 until 10142295. The third line is the 2nd incremental job with ID 338.

To further prepare the restored files, use the xtrabackup --prepare command. Read https://www.percona.com/doc/percona-xtrabackup/2.4/xtrabackup_bin/incremental_backups.html for more information.

Backup of MySQL Databases using the Python MySQL plugin

The Python plugin from https://github.com/bareos/bareos-contrib/tree/master/fd-plugins/mysql-python makes a backup of all or selected MySQL databases from the Bareos File Daemon or any other MySQL server. It makes use of the mysqldump command and basically grabs data from mysqldump via pipe. This plugin is suitable to backup database dumps. If you prefer to use mechanisms like incremental hot-backups of InnoDB tables, please use the Bareos MySQL / MariaDB Percona xtrabackup Plugin (see Backup of MySQL Databases using the Bareos MySQL Percona xtrabackup Plugin).

Following settings must be done on the Bareos client (Bareos File Daemon):

  • install and enable the Bareos File Daemon Python plugin
  • install the Python MySQL plugin (for some platforms it is available prepackaged from http://download.bareos.org/bareos/contrib/, on the other platforms: copy the plugin files to the Bareos Plugin Directory)
  • disable bacula compatibility (default for Bareos >= 15.2)
bareos-fd.d/client/myself.conf
Client {
  ...
  Plugin Directory = /usr/lib64/bareos/plugins
  Plugin Names = "python"
  compatible = no
}

Configure the plugin in the Bareos Director:

bareos-dir.d/fileset/mysql.conf
FileSet {
    Name = "mysql"
    Include {
      Options {
        signature = MD5
        compression = lz4
      }
      Plugin = "python:module_path=/usr/lib64/bareos/plugins:module_name=bareos-fd-mysql:db=test,wikidb"
      #Plugin = "python:module_path=/usr/lib64/bareos/plugins:module_name=bareos-fd-mysql:mysqlhost=dbhost:mysqluser=bareos:mysqlpassword=bareos"
    }
}

In the above example the plugin creates and saves a dump from the databases called test and wikidb, running on the file-daemon. The commented example below specifies an explicit MySQL server called dbhost, and connects with user bareos, password bareos, to create and save a backup of all databases.

The plugin creates a pipe internally, thus no extra space on disk is needed. You will find one file per database in the backups in the virtual directory /_mysqlbackups_.

List of supported options:

db
comma separated list of databases to save, where each database will be stored in a separate file. If ommited, all databases will be saved.
dumpbinary
command (with or without full path) to create the dumps. Default: mysqldump
dumpoptions
options for dumpbinary, default: “–events –single-transaction
drop_and_recreate
if not set to false, adds –add-drop-database –databases to dumpoptions
mysqlhost
MySQL host to connect to, default: localhost
mysqluser
MySQL user. Default: unset, the user running the file-daemon will be used (usually root)
mysqlpassword
MySQL password. Default: unset (better use my.cnf to store passwords)

On restore, the database dumps are restored to the subdirectory _mysqlbackups_ in the restore path. The database restore must be triggered manually (mysql < _mysqlbackups_/DATABASENAME.sql).

Backup of a MySQL Database by using the RunScript directive

One method to backup a MySQL database is to use the mysqldump tool to dump the database into a file and then backup it as a normal file. After the backup, the file can be removed. It may also be an option not to remove it, so that the latest version is always available immediately. On the next job run it will be overwritten anyway.

This can be done by using Run Script (Dir->Job) directives, for example:

RunScript job resource for a MySQL backup
Job {
  Name = "BackupDatabase"
  JobDefs = "DefaultJob"
  Client = dbserver-fd
  Level = Full
  FileSet="Database"

  # This creates a dump of our database in the local filesystem on the Client
  RunScript {
    FailJobOnError = Yes
    RunsOnClient = Yes
    RunsWhen = Before
    Command = "sh -c 'mysqldump --user=<username> --password=<password> --opt --all-databases > /var/lib/bareos/mysql_dump.sql'"
  }

  # This deletes the dump in the local filesystem on the Client
  RunScript {
    RunsOnSuccess = Yes
    RunsOnClient = Yes
    RunsWhen = After
    Command = "rm /var/lib/bareos/mysql_dump.sql"
  }
}

FileSet {
  Name = "Database"
  Include {
    Options {
      signature = MD5
      compression = gzip
    }
  # database dump file
  File = "/var/lib/bareos/mysql_dump.sql"
  }
}

Note that redirecting the mysqldump output to a file requires to run the whole command line through a shell, otherwise the mysqldump would not know what do with the > character and the job would fail. As no shell features like redirection or piping are used for the rm, the sh -c is not needed there. See Run Script (Dir->Job) for more details.

Backup of a MySQL Database by using the bpipe plugin

Instead of creating a temporary database dump file, the bpipe plugin can be used. For general information about bpipe, see the bpipe Plugin section. The bpipe plugin is configured inside the Include section of a File Set, e.g.:

bpipe fileset for MySQL backup
FileSet {
  Name = "mysql-all"
  Include {
    Plugin = "bpipe:file=/MYSQL/dump.sql:reader=mysqldump --user=<user> --password=<password> --opt --all-databases:writer=mysql --user=<user> --password=<password>"
    Options {
      signature = MD5
      compression = gzip
    }
  }
}

This can also be used, to backup a database that is running on a remote host:

bpipe directive to backup a MySQL database that is running on a remote host
FileSet{
  Name = "mysql-all"
  Include {
    Plugin = "bpipe:file=/MYSQL/dump.sql:reader=mysqldump --host=<hostname> --user=<user> --password=<password> --opt --all-databases:writer=mysql --host=<hostname> --user=<user> --password=<password>"
    Options {
      signature = MD5
      compression = gzip
    }
  }
}

If you do not want a direct restore of your data in your plugin directive, as shown in the examples above, there is the possibility to restore the dump to the filesystem first, which offers you more control over the restore process, e.g.:

bpipe directive to backup a MySQL database and restore the dump to the filesystem first
FileSet{
  Name = "mysql-all"
  Include {
    Plugin = "bpipe:file=/MYSQL/dump.sql:reader=mysqldump --host=<hostname> --user=<user> --password=<password> --opt --all-databases:writer=/usr/lib/bareos/scripts/bpipe-restore.sh"
    Options {
      signature = MD5
      compression = gzip
    }
  }
}

A very simple corresponding shell script (bpipe-restore.sh) to the method above might look like the following one:

bpipe shell script for a restore to filesystem
#!/bin/bash
cat - > /tmp/dump.sql
exit 0

Statistics Collection

Statistics Collection can be controlled by a number of configuration directives. If Statistics Collection is enabled, statistics are collected by the Bareos Director and stored into the Catalog database. So enabling this feature will increase your database size.

The Statistics are used by the Bareos WebUI to show the status of a running job.

Director Configuration - Director Resource Directives

Director Configuration - Storage Resource Directives

Storage Configuration - Device Resource Directives

See chapter Job Statistics for additional information.

Removing a Client

Removing a client can mean several things in Bareos context. You can disable the client so no jobs that use this client will run anymore. You can also remove the client’s configuration so that no job can use the client anymore. However, you will still have the client in your catalog unless you explicitly remove it.

Disable a client

It is possible to disable a client either by calling disable client=<client-name> or by setting Enabled (Dir->Client) to no. While the configuration change will persist across restarts, the command just disables the client temporarily.

When a client is disabled, the scheduler will not run any job for this client anymore, but you can still restore to it or manually trigger a job.

Remove configuration for client

If you want to remove the client permanently, you can also completely remove the client from your configuration. This will make sure that no job (neither backup nor restore) can use that client machine. After removing the Client Resource from the director configuration, you will also need to change or remove all Job and JobDefs Resources in the Director configuration that reference the removed client.

Removing the client configuration does not remove the backed up data or the file and job records in the catalog for the client. You can still restore the data that has been backed up from the client to a different client.