Using a snapshot of a replica for mailbox database recovery in Database Active Group (DAG)

A snapshot of a replica can be used to restore an Exchange mailbox database to the point of failure (POF) or to a point in time (PIT). Taking a snapshot of a replica can be done through the VEA GUI by using either the VSS snapshot wizard or with the vxsnap CLI command. Taking a snapshot of a replica automatically takes a snapshot of all the volumes that the replica uses.

When SFW fails to automatically dismount the database during a restore operation, the restore operation fails. The restore operation can be performed again after manually dismounting the databases and manually setting the databases for overwrite by restore.

Although SFW lets you restore the database from a snapshot, restoring just the database log files is not supported.

For an Exchange 2010 database that is configured under VCS cluster, offline the corresponding VCS database resource before the restore operation if automatic offlining of resource database fails. SFW automatically offlines the resource databases. Once the restore operation completes successfully, online the VCS resource database.

In other configurations, SFW dismounts the databases before the restore operation and then mounts the databases after the restore operation automatically.

Using the -a option automatically dismounts and mounts the databases. For a VCS cluster setup, it offlines and onlines the resource databases.

Refer to Microsoft Exchange 2010 Shell Help for details on Exchange cmdlets.

An example of a PIT recovery procedure from a snapshot of a replica copy of an Exchange mailbox database DB1 on an active mailbox database TestServer1and Replica copy on TestServer2 would be as follows: