Description

NOTE: This cookbook is forked from the awesome work done by computerlyrik to add additional features (e.g. tape drive support). Its name has been changed to bacula-backup to allow for a unique cookbook name to be used on the Supermarket. It may be renamed to just bacula again in the future.

Tested on Ubuntu 12.04 server (should also work on Debian - please report!). Also supports file daemon (e.g. 'client') installs on RHEL 5 & Windows.

Supports File backup on a dedicated storage server with a single bacula director and multiple clients to be backed up.

Painlessly autoconfigures itself and the Jobs to be done.
Backs up daily incremental, full weekly. (see usage)
Also supports installation of bacula console to observe your backup progresses.

Autogenerated Jobs

• Files
• LDAP
• Chef server

Recipes

Bacula Director (bacula-dir)

bacula-backup::server 

Central backup server

Bacula File Daemon (bacula-fd)

bacula-backup::client

Used by each client to be backed up. Currently supports Ubuntu 12.04/14.04, RHEL/CentOS 5, and MS Windows.

Bacula Storage Daemon (bacula-sd)

bacula-backup::storage (bacula-sd)

For use on storage system

Bacula Administration Tool BAT (bacula-console-qt)

bacula-backup::bat

Used for Systems with graphic environment - installs and configures "bat" Bacula qt-console

Requirements

See metadata for cookbook dependencies.

Attributes

default.rb

Configure the bacula user
ruby
node['bacula']['user']
node['bacula']['group']


server.rb

Set general backup properties
ruby
node['bacula']['dir']['max_concurrent_jobs'] = "1"
node['bacula']['dir']['address'] = nil # if set, causes to director to bind to the specified address


Set properties for File based backup
ruby
node['bacula']['volume_size'] = "1G"
node['bacula']['volume_max'] = 20
node['bacula']['label_format'] = "BaculaFile"


Additional storage pools (beyond the built-in File & Default pools) can be specified as arrays in the ['bacula']['dir']['pools'] attribute. E.g.:
ruby
default['bacula']['dir']['pools'] = [
{
"Name": "MyTapePool",
"Volume Retention": "365 days",
"Storage": "\"mystorage.local-mytapelibrary-LTO-5\"",
"Pool Type": "Backup",
"RecyclePool": "Scratch",
"Recycle Oldest Volume": "yes"
}
]

The contents of this array will be directly converted into Pool resources in the storage configuration file. E.g., the above example would be translated to:

Pool {
Name = MyTapePool
Volume Retention = 365 days
Storage = "mystorage.local-mytapelibrary-LTO-5"
Pool Type = Backup
RecyclePool = Scratch
Recycle Oldest Volume = yes
}


client.rb

Set files to be backed up (see Usage below)
ruby
node['bacula']['fd']['files']


storage.rb

Set up destination of File-Storage
ruby
default['bacula']['sd']['backup_dir'] = "/backup"


Configures the Maximum Network Buffer Size for the File-Storage device
ruby
default['bacula']['sd']['file_max_net_buffer'] = "32768"


Additional storage devices (e.g. tape devices and autochangers) can be specified in the ['bacula']['sd']['devices'] or ['bacula']['sd']['autochangers'] attributes (and then again in ['bacula']['sd']['dir-devices'] -- see below). E.g.:
ruby
default['bacula']['sd']['devices'] = [
{
"Name": "LTO5-1",
"Media Type": "LTO-5",
"RemovableMedia": "yes",
"RandomAccess": "no",
"Archive Device": "/dev/nst1"
}
]

The contents of these arrays will be directly converted into Device or Autochanger resources in the storage configuration file. E.g., the above example would be translated to:

Device {
Name = LTO5-1
Media Type = LTO-5
RemovableMedia = yes
RandomAccess = no
Archive Device = /dev/nst1
}

Because options for the corresponding director Storage resources can vary (e.g., are devices in an autochanger? If yes, don't make their own Storage resource), these need to be manually specified. E.g.:
ruby
default['bacula']['sd']['dir-devices'] = [
{
"Device": "mytapelibrary",
"Media Type": "LTO-5",
"AllowCompression": "no",
"Autochanger": "yes"
}
]

Note that the Name, Address, SDPort, and Password attributes must be ommitted as they are automatically configured. The contents of this array will be directly converted into Storage resources in the director configuration file. E.g., the above example would be translated to the following (assuming a server with a FQDN of mystorage.local, ip of 1.2.3.4, and password of 12345):

Storage {
Name = mystorage.local-mytapelibrary-LTO-5
Address = 1.2.3.4
SDPort = 9103
Password = "12345"
Device = mytapelibrary
Media Type = LTO-6
AllowCompression = no
Autochanger = yes
}


Usage

Backup Job generation

To autogenerate jobs the following expressions need to be true on bacula-backup::client machine:

Ldap

node['openldap'] && node['openldap']['slapd_type'] == "master"

Chef Server

node['fqdn'] == "chef.#{node['domain']}"

Default deployment

• node A => bacula-backup::server
• node B => bacula-backup::storage (with much storage)
• node C-Z => bacula-backup::client

Examples

Simple file backup

Set on your bacula-backup::client node
ruby
node.set['bacula']['fd']['files'] = {
'includes' => ['/']],
'excludes' => [ '/dev','sys']
}


Custom Backups

When more granual control over backup jobs/filesets/schedules is needed, they can be specified as arrays on the clients' ['bacula']['fd']['jobs'], ['bacula']['fd']['filesets'], & ['bacula']['fd']['schedules'] attributes. E.g.:

"default_attributes": {
  "bacula": {
    "fd": {
      "filesets": [
        [
          "Name = \"myclient-accurate-opt\"",
          "Ignore FileSet Changes = yes",
          "Include {",
          "  Options {",
          "    signature = MD5",
          "  }",
          "File = /opt",
          "}"
        ]
      ],
      "jobs": [
        {
          "Name": "\"myclient-wednesdays\"",
          "Client": "myclient.local",
          "Type": "Backup",
          "Accurate": "yes",
          "Level": "Incremental",
          "Messages": "Standard",
          "FileSet": "\"myclient-accurate-opt\"",
          "Schedule": "\"myclient-wednesday\"",
          "Pool": "Default",
          "Storage": "File"
        }
      ],
      "schedules": [
        [
          "Name = \"myclient-wednesday\"",
          "Run = Level=Full 1st wed at 18:00",
          "Run = Level=Incremental 2nd wed at 18:00",
          "Run = Level=Incremental 3rd wed at 18:00",
          "Run = Level=Incremental 4th wed at 18:00",
          "Run = Level=Incremental 5th wed at 18:00"
        ]
      ]
    }
  }
}

Note that schedules and filesets are specified as raw arrays under their parent array (because of their varied and often complex structure), and jobs are specified are hashes under their parent array. The above example will generate the following director entries:

FileSet {
  Name = "myclient-accurate-opt"
  Ignore FileSet Changes = yes
  Include {
    Options {
      signature = MD5
    }
  File = /opt
  }
}

Job {
  Name = "myclient-wednesdays"
  Client = myclient.local
  Type = Backup
  Accurate = yes
  Level = Incremental
  Messages = Standard
  FileSet = "myclient-accurate-opt"
  Schedule = "myclient-wednesday"
  Pool = Default
  Storage = File
}

Schedule {
  Name = "myclient-wednesday"
  Run = Level=Full 1st wed at 18:00
  Run = Level=Incremental 2nd wed at 18:00
  Run = Level=Incremental 3rd wed at 18:00
  Run = Level=Incremental 4th wed at 18:00
  Run = Level=Incremental 5th wed at 18:00
}

Contact

see metadata.rb