Hi HeCSa, OpenSVC synchronize datasets, relying on ZFS snapshot feature. Snapshots are recursive, so using one pool to store all postgresql database ensure an atomic point in time recovery. So service may be restarted on other nodes as it was at the time of snapshot (if PostgreSQL database allow recover after OS crash it should work).
You may declare a service as describe in attached file, add startup scripts into /opt/opensvc/etc/mydatabase.d/ S10postgresql K90postgresql C50postgresql Then each time you launch each time you want to refresh node2 and node3 Launch from r...@node1: /opt/opensvc/etc/postgresqldb syncupdate You will have to reserve space on ZFS pool (space to keep "delta" between live and last synchronisation) See also: http://www.opensvc.com/init/default/doc_storage_zfs Best regards Cyril -- This message posted from opensolaris.org
example
Description: Binary data
_______________________________________________ sysadmin-discuss mailing list sysadmin-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/sysadmin-discuss
