Unless there's some positive indication that the array is functioning improperly, I'm not sure what NetApp support can do for you here. There's no real way for the storage array to call eseutil (unless there's some sort of confusion where DPM is calling the NTAP hardware provider - an implausible scenario during a simple node failure). DPM is the application calling for verification, and is either incorrectly recognizing an MSCS failover as an event that necessitates a verification, or distrusts the integrity of the databases after a failover. So it's either a bug or the intended result of a failover.

If I were the one responsible for opening (not closing) the helpdesk ticket, I would ask that Microsoft either fix the problem, give me a workaround, or provide definitive proof that the problem resides with the storage (i.e., the array is changing the state or content of NTFS or the exchange DBs or logs during a failover). Only if they provided the last would I open a ticket with NetApp...

On Mon, Nov 24, 2008 at 12:35 PM, Darish Rajanayagam <darishr@softcom.biz> wrote:

Toasters,

Does anyone use Microsoft DPM for Exchange backups in a MSCS cluster using NetApp storage(iscsi)? Whenever the Exchange virtual server is moved over to a new node, DPM wants to perform consistency checks on the protected storage groups – according to Microsoft this should not be the case – if I were to expand the volume, we would need to run DPM consistency checks but not during normal cluster failover (DPM performs block level backup).

I am having an issue where after every failover, I have to re-run consistency checks on all my storage groups (100 Storage groups!). I had a case open with Microsoft who asked me to contact NetApp – NetApp support wasn't very helpful in this matter. Before I proceed further, I am wondering if anyone else is experiencing this issue.

Thanks,

Darish