I recently implemented Veeam (implemented v5.0.2 the day before 6 was released, wanted a stable release). Veeam is installed on a single 2008 x64 Server machine and all works great except for one VM. I'm running all of my jobs over direct SAN access (Dell Equallogic PS5000e iSCSI SAN). All of my VMs work fine except for one SQL VM that hangs at 0%, but never errors or fails, it will just sit at 0%. In troubleshooting this, I've isolated the problem to a specific disk that is an RDM that was created in virtual compatibilty mode. This RDM disk contains the SQL databases for the server. I've confirmed that Veeam creates the snapshots of both the VMFS volume disk and the RDM when I launch the job. The job then sits at 0% (I've let it sit overnight) but it never starts copying the VMDK data over. I then attempt to stop the job, which never works, am forced to kill the agent.exe and manager.exe processes, and remove the snapshot from snapshot manager. This process of removing the Veeam snapshot doesn't actually remove the snapshot files, they're still present in the VMFS volume where the VMDK files live, but the snapshot is no longer listed in the Snapshot Manager. I then have to manually create another snapshot and then delete all to get the Veeam snapshot to go away...
Physical Server 2008 x64 Veeam server
13x 500GB SAS drives in Raid 60
2x Intel Xeon E5649
16x 500GB Raid 5 Dell Equallogic PS5000e iSCSI SAN
CPU and memory appear normal while the job is hung
NIC activity bounces between 0% and 0.5% activity- I'm pretty sure the only traffic is SOAP requests to keep the snapshot alive.
This job works fine
in Network access mode
All other VMs process fine over direct SAN access mode
I can manually create snapshots for the RDM through VSphere
The SQL server isn't overly fragmented
Would really like to get direct SAN access working on this box so I can back it up during business hours w/o flooding the front end network.