Exchange Backup Behavior

[tjackson111]

Stumbled across what seams to be an odd issue backing up an Exchange DAG. Two servers with four databases, one database per volume. Each server has two active and two passive copies. My intent was to use the agent on both servers and selecting only the volumes that housed the passive copies. However, what I am finding is that Veeam never tells Exchange that the DB was backed up and to truncate the logs. The logs seem to imply that even though Exchange was detected, because I only selected two of four volumes with Exchange data, that it did not enable the Exchange VSS writer and excluded it from the job. I confirmed this by selecting all Exchange data volumes and it worked fine. I would think I should be able to select only certain Exchange DB's for backup right? Is this a bug or some kind of feature by design?

Thanks,
Tim

[Dima P.]

Hello Tim,

Unfortunately, failover clusters and residing applications are not supported in current version on Veeam Agent. Take a look at this KB for more details. Regarding your issue, to make truncation work all databases should be included in the backup (so if some volumes are excluded exchange most likely won’t truncate logs).

[JWester]

Hello Dima,
do you know if this will be fixed in the next release? We have the same problem on two DAG-clusters. And backing up twice the data needed is not the optimal solutions.
Many thanks!

[vClintWyckoff]

The upcoming release of Backup & Replication 9.5 Update 3 and VAW 2.1 will support Exchange DAG on bare-metal hardware. Just as an FYI MS Exchange DAG does not support concurrent VSS snapshots on multiple DAG nodes so the new cluster aware backup job will fail. As a workaround, similar to our VBR recommendation is to use the regular VAW server backup job to protect a single DAG node. If you have multiple active DAG servers in the environment you can use the VBR job chaining capability to start backup of another DAG node after the first completes.

[Andreas Neufert]

By design of Exchange it is anyway not best practice to backup the Exchange DAG member Servers at same time. If something goes wrong and the Cluster failover to the other DAG server it could happen that the other server failover then at nearly same time which can lead massive CPU usage from several services like index service and can lead into timouts at outlook and other services (POP/IMAP/OWA).

Will update my best practices after VAW2.1 goes online... http://andyandthevms.com/exchange-dag-v ... plication/