Frequently Asked Questions

• Why do we have to specify the --stanza option even if only one stanza is defined in the configuration file?

Except for the repository specific commands (info, start, stop), the --stanza option is required so that adding a stanza does not cause the commands that did not have a stanza specified to suddenly throw errors because they do not know which stanza to use.

• Is it possible to only restore a specific database from a complete database cluster backup?

Yes. You can restore a specific database using the restore --db-include option. After recovery, you can then extract the selected databases with the pg_dump command.

The databases that are not explicitly included with this option will be restored as sparse, zeroed files to save space; however, it still allows PostgreSQL to perform the recovery using less disk space. You can then remove these databases with the DROP DATABASE command since they will no longer be accessible.

• Is is possible to setup pgBackRest to access remote hosts without using the SSH protocol?

Yes. Since the 2.37 release, pgBackRest can use TLS with client certificates to enable communication between the hosts. A TLS server must then be configured and started on each host.

See [Setup TLS[(https://pgbackrest.org/user-guide-rhel.html#repo-host/config) for more information.

• How many CPU cores are needed on a dedicated remote backup server?

Since compression/decompression is performed on the Postgres host, the repository host does only need to encrypt/decrypt SSH or TLS connections. Usually 1 repository core per 4 Postgres cores is a good starting point.