Hi, On Thu, Apr 2, 2015 at 4:32 PM, Denish Patel <denish@xxxxxxxxxx> wrote: > Alexey, > > Instead of managing it manually, you should be able to take backup from > slave using OmniPITR: > https://github.com/omniti-labs/omnipitr/blob/master/doc/omnipitr-backup-slave.pod Thanks for the suggestion. We don't manage replica creation manually, we have a tool that is just not designed to get base backups from the standby, because we never needed it until recently. Instead of adopting another tool I'd rather understand the process and implement the necessary changes in the tool of our own. So far, there are 2 solutions (assuming archiving from the master is turned on and archives are available to the new replica): - run pg_start_backup from the master, wait until the replica reaches the same position as a master during the backup, copy the data files from the replica, copy the backup label from the master, run pg_stop_backup and start the new replica. - rely on pgespresso extension and run pgespresso_start_backup/pgespresso_stop_backup on the replica (and grab a backup label in between), and do the rest of the process the same way like it's done when taking base backups from the master. And, just for completeness, a set of wonderful tools to help: - pg_basebackup (once has to turn on full_page_writes on the master beforehand) - barman - omnipitr Thank you, Alexey. -- Sent via pgsql-admin mailing list (pgsql-admin@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-admin
