I'm managing a project to implement netbackup 6 on a 3-node clusters environment where clusters node is running on public network while netbackup on the backup network. Netbackup is setup with Rman and there is 2 applications databases sitting on node-1 together with rcat (rman). Each DB have one public VIP (for normal operations) and one backup VIP (for backup use only) assigned. Node-3 on the cluster is used for failover, during failover, there are possibilities that any of the database will be failover to node-3 independently, creating situation where all DBs are not colocated on one node. Node2 is not used for DB.
It is a known limitation that the NBU configuration file can only support one VIP so as a workaround, we have created multiple backup policies, each with a different VIP, scheduled to run at diffent time, one at a time.
Backup test has so far successful when both application databases is co-located on same server (node-1 or node-3) regardless of where rcat is but failed when both application databases are not co-located (either DB1 on node 1 and DB2 on node3 or DB1 on node3 and DB2 on node1).
Our troubleshooting (with Symantec support) has so far reveal no clue as to where the problem is. If anybody has experienced the same problem before or have any idea on where it might went wrong, please kindly help!
Thanks a million
