VSS timeout - Exchange 22 disks

[b.vanhaastrecht]

Hi,

I'm facing a problem with an very large Exchange server farm, Veeam is giving a VSS timeout problem. I'm folowing this KB in an attempt to resolve the issue: http://www.veeam.com/kb1744

I have some problems interpeting solution number 3 (Perform the Exchange Backup singularly as opposed to concurrently with other jobs.). Does this mean parralel processing must be disabled for this job? To my knowledge this is not a per job setting. Or is there another interpetation?

kind Regards,
Bastiaan

[foggy]

Bastiaan, I guess what is meant there is that Exchange VM backup should be performed as fast as possible, so it is recommended to back it up at the time when there are no other jobs running, so that all the available resources could be devoted exclusively to this job. You should not disable parallel processing, as it allows to process VM disks in parallel thus reducing the time it lives on snapshot.

[b.vanhaastrecht]

Ok, understand. We are running the job seperate from all others, so thats checked.

Solution number 6 refers to reducing the amount of disks. The Exchange database servers have 22 disks, which is quite large. But before we start a consolidation we need to be sure this solution Is going to help in a fair amount. Is there some way to determen to find the sweet spot on the amount of disks that could stop the VSS timeout error to occour?

[Vitaliy S.]

Less virtual disks you have, the less time it should be required to take a VSS snapshot and then commit the data. I'm not sure there is a recommendation that would allow you to decide on the optimal number of virtual disks, but you may want to look through this topic (if you haven't done that already) for some other tips from Andreas regarding exchange backups > VSS timeout with Exchange 2010

[b.vanhaastrecht]

Because the amount of disks, 22, was at a high rate we've consolidated this to 8. After that we've shutdown the VM, updated the VMware tools (missed a update), and added 8GB RAM to the VM because RAM was under pressure the application aware backup went successfully.