Bareos - Backup Archiving REcovery Open Sourced
Bareos is a fork of the popular open source backup solution, Bacula. Aside from improvements such as the default assignment of environment variables, Bareos adds features that are not found in Bacula.
Backup and restore
As you would expect, Bareos covers all normal backup system functions:
- Scheduling of backup and job verification according to individual timetables
- Full, incremental and differential backups
- Data compression before storage
- Concurrent job prioritization
- Restoration of individual files or a complete client
Extensive network integration
Via a backup client, all relevant operating systems (Linux, Windows, Unix, Macintosh, etc.) can be backed up remote from the server. Authentication between server and client is secured with a CRAM-MD5 encrypted password.
Because a system backup requires only one TCP port which can be released by firewalls, backups are also possible through a firewall. To save on bandwidth and memory, data is compressed up to 70% before transmission. Scripts can be run on the client before and after backups, eg. taking an application offline or executing a database dump.
Bareos introduces an additional feature that allows the file daemon to be placed in a DMZ and provided with relevant backup files. Both the organisation of, and communication between agents around the Bareos server have been adapted, such that: The moment the file daemon receives a notification, updates are run and the client is notified that the backup process may begin.
This feature is explained in the following diagram:
Database catalogue management
All system transaction data is stored and archived on a local MySQL database. This allows individual jobs, different files or complete clients be restored. It is even possible to target database files matching a certain date to restore. With an individual expiration date, these entries are again later removed from the database.
The backed up data, compressed or uncompressed are saved in a platform independent format so both Linux and Windows backups can be saved in the same archive file. If the backup server (in a worst case scenario) was unavailable, then data can be directly extracted from the memory files through low-level tools.
For normal daily administration, the console based front end offers a lean and fast management interface. Generated logs can also be viewed on this interface or sent to an administrator, depending on its configuration and how much information you want to receive.