Error Log in Microsoft Hyper V Shared-VHDX section after backup
Hi All
I have a brand new 2012 R2 cluster with 4 nodes, all verification check complete with success and there is no visible problems with the vm or backup operation on the cluster.... however when i was scanning through the application and service logs on the host servers i came across the below error logs.
Error attaching to volume. Volume: \Device\HarddiskVolumeShad
this error occurs after every backup and appears on the host server that has current ownership of the csv that the vm is housed on when it is backed up.
I have carried out a manually disk shadow and this also pulls up the same error, so this rules out my current backup provider. What is strange is that the error is listing in the shared vhdx section but the vm's that being backed up are vhd's not vhdx.
Has anyone else had this problem?
Thanks
Daniel
I have a brand new 2012 R2 cluster with 4 nodes, all verification check complete with success and there is no visible problems with the vm or backup operation on the cluster.... however when i was scanning through the application and service logs on the host servers i came across the below error logs.
Error attaching to volume. Volume: \Device\HarddiskVolumeShad
this error occurs after every backup and appears on the host server that has current ownership of the csv that the vm is housed on when it is backed up.
I have carried out a manually disk shadow and this also pulls up the same error, so this rules out my current backup provider. What is strange is that the error is listing in the shared vhdx section but the vm's that being backed up are vhd's not vhdx.
Has anyone else had this problem?
Thanks
Daniel
ASKER
Backup product is veeam
i haved moved the vm to the same node as the storage and I have run a manual snapshot and interestly i have not got the error logged this time??
i haved moved the vm to the same node as the storage and I have run a manual snapshot and interestly i have not got the error logged this time??
Correct. CSV redirection is what happens when the storage owner and VM host are different.
Veeam should have some documentation and possibly the PowerShell to move things into sync.
Veeam should have some documentation and possibly the PowerShell to move things into sync.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
The only way to "prove" that is to do a full bare metal or hypervisor restore of the backups.
EDIT: Hang on a sec, the errors go away when the CSV and VM are on the same owner but come back when they are not? I'm sorry to say it, but that is _not_ some sort of transient error due to a bug in the Event Viewer. We see this situation in many cluster settings where I/O issues happen when CSV and VM are on different owners. It manifests itself in many ways!
EDIT: Hang on a sec, the errors go away when the CSV and VM are on the same owner but come back when they are not? I'm sorry to say it, but that is _not_ some sort of transient error due to a bug in the Event Viewer. We see this situation in many cluster settings where I/O issues happen when CSV and VM are on different owners. It manifests itself in many ways!
ASKER
Hi Phillip,
After further testing it makes no difference whether the csv and vm are on the same node i still got the error in the logs when backing up or taking a manual disk shadow. I did not get the error when i carried out a checkpoint in hyper vmm however this is not the same as a disk shadow. I have been able to restore backups of csv and vm without any problem too. MS have checked the cluster config and everything is how it should be.
Thanks for your help
After further testing it makes no difference whether the csv and vm are on the same node i still got the error in the logs when backing up or taking a manual disk shadow. I did not get the error when i carried out a checkpoint in hyper vmm however this is not the same as a disk shadow. I have been able to restore backups of csv and vm without any problem too. MS have checked the cluster config and everything is how it should be.
Thanks for your help
ASKER
Thanks for your help, i have contacted MS and they have advised that the logging bears no issue and this is a bug problem.
Try running the manual snapshot while the VM is on the storage owner.
There are PowerShell scripts out there that move storage ownership to the same node as the VM to keep them together. Otherwise, I/Os are redirected with the requisite issues that can come along with that.