Networker Module for Microsoft Applications 2.3.1 (2.3.0.build81) is really doing well backing up Microsoft Exchange 2010 DAG configurations.
But a few weeks ago we ran into some strange issue.
Suddenly the Networker group backup was unsuccessful as you can see in the screenshot below, failing with the messages:
“NMM Backup failed to complete successfully” and “nsrsnap_vss_save createInstanceBackup() failed”
After some investigation, we noticed that the other node was doing fine.
And then we saw that all Exchange 2010 databases were mounted on 1 node, the other node without mounted/active databases was having problem.
I posted this on the EMC community and an EMC support engineer commented that this was normal behaviour: “NMM design is if there is not respective backup type mailbox database found while running DAG backup, then we will fail backup.”
So if you have a cluster and for some reason, can’t think of one, but for some reason all databases are mounted on only 1 of the nodes, then NMM will run into problems when trying to backup the “passive” node.
You will have 2 options:
1. Mount a database on the other node
2. disable the Networker Client Resource for the passive exchange node and change for the active exchange node, the networker parameter “NSR_EXCH2010_BACKUP=active -> NSR_EXCH2010_BACKUP=all”.
To me… the above is a dirty-workaround… I would really encourage EMC² to add an extra check into NMM that will not produce this failure, but warn the user in stead.