Hyper V Host Lost CSV (c000020c) while Backup

[max0664]

Hi,

I have a Hyper V Cluster with 6 Nodes and CSV Disks (Fiber Channel). 2 Nics for Guest only and 2 Nics for all other traffic.
OS: Windows Server 2008 R2
AV: ForeFront with exclude ClusterStorage

I do Backups with Veeam. But sometimes i get this Error while the backup is running:

Code: Select all

Cluster Shared Volume 'Volume2' ('Cluster Disk 2') is no longer available on this node because of 'STATUS_CONNECTION_DISCONNECTED(c000020c)'. All I/O will temporarily be queued until a path to the volume is reestablished.

The Guest on the this note with the disk on the Cluster disk 2 are Restarting.

Someone have idea why my Note lost the Connection to the CSV?

[Gostev]

Hi Max, what is your support case number for this issue? Thanks.

[max0664]

i dont have a case number i only tried veeam

[Gostev]

If you'd like this investigated for you, please open a support case. I cannot really help you until you do this, because this will be the first thing developers will ask me about when I approach them with this. They always want to see debug logs first...

[jonhsykes]

I realize this is an old thread however I have this same issue. A support case has was opened yesterday (02124784); unfortunately no advances have been made.

[foggy]

Hi Jon, I'm not able to check for the solution, since the OP didn't provide his support case ID. Please continue working with your engineer.

[jonhsykes]

Hi foggy, unfortunately the engineer that emailed me yesterday didn't fully "engage" in the problem yet. This is what I received, response in bold.

"Hello Jon,

Thank you for contacting Veeam Support. The logs show problems with C:\ClusterStorage\Volume19 and C:\ClusterStorage\Volume20 - these two locations are a part of most of the errors that we are seeing at the moment. That is correct. Once the replication for these CSVs/VMs on these CSVs failed, the host was unresponsive, cluster service stopped, VMs were in a failed state.

Are these two parts of the storage different from the rest of the clustered storage ? Can you tell me the hardware details of the storage and how it is connected ? Can you rescan the host and retry this replication job ? This is Hitachi VSP storage, using 2x Fibre Channel HBAs with MPIO. It is the same as other Hyper-V clusters in the environment, some of which we have replicated successfully. Since the replication job caused the cluster node/hyper-v host to fail I would prefer not to kick the job off yet until we determine root cause or at least have a better understanding of what happened so we can proceed forward.

Please let me know what happens next."