Backup Freeze on VM after Storage vMotion

[lobo519]

I moved a VM to a different datastore last night (Storage vMotion) and when my backup job ran it backed up a few VMs and then froze on the one I moved for several hours before I just cancelled it. But it wouldn't cancel. I tried restarting the Veeam services with no luck so I finally just rebooted the Veeam server which cleared the job.

Anyone have any issues like this?

[ian0x0r]

Ive just had this exact same issue, on a replication job. Although restarting the job after server reboot, it has hung again.

[ian0x0r]

Actually, I tell a lie, the job has started again after a reboot to the server. Its just running slowly for some reason.

[Libertas]

Hi, Have you tried the same again, starting a backup and then performing a Storage vMotion.

I had the same problem that when I started a backup job, for example with 10 VMs, and any of the VMs that are not being backed up at that time where Storage vMotioned before the backup job got round to backing them up, the backup job would stick at "building VM list".

I never managed to sort it out with support, so It would be interesting to see if it was just me with the problem.

Again the scenario, start a backup job and then mid way in that job a VM gets Storage vMotioned before the backup job starts to back up that VM. Once the job get to the point were it wants to backup the moved VM, it stalls with "Building VM List" and does not go further.

Thanks,

[cby]

Hi

Did anyone find a resolution to this problem?

I had the same issue last night after migrating a set of hard disks on a VM to a new datastore. The odd thing is that the backup jobs that hung with 'Building VMs list' message were not part of the migration nor did they start the backup until after the migration had finished. Very odd.

The only way to cancel the hung jobs is to kill the Veeam agent processes.

I haven't rebooted the Veeam server.

[cby]

Update: Reboot Veeam server made no difference.

[rbrambley]

Running a storage VMotion or even just a VMotion should probably be avoided during (anybody's) backup jobs in general. Both can be IO and CPU intensive operations, and depending on the rest of VI architecture could cause an impact that is exponentially greater during any kind of backup job. Not to mention if the Veeam server is also a VM, using the same datatstores and hosts of the VMs it is backing up. Ultimately, you probably should work with Veeam Support to get to the bottom of this, but here are some other things to consider:

How did you configure the backup job? Did you select each VM by name or did you pick a parent object like a host, folder, resource pool, LUN, etc? Is your change / SVMotion moving the VM out of the previous job configuration scope?

How did you add the hosts to the Veeam Server? Did you add individual hosts or add vCenter? vCenter is obviously aware of Clusters, shared LUNs, DRS, etc. Adding hosts individually does not account for dynamic cluster level operations.

If your job configs allows it, remove the VM that is freezing from the job. Save the job, and then add the VM back again. Veeam has a back end SQL database that needs to update.

Watch your vSphere environment performance and SAN during backups, VMotions, and SVMotions. try to see if the CPU / Mem/ and IO generated is abnormal, and if so, can the underlying hardware (hosts, disk arrays) handle it? Use Veeam Monitor for this!

There is a lot more to consider that could be the cause of this issue then just Veeam BaR.

[cby]

Rich

Thanks for the detailed response.

Things have moved on, some would say in mysterious ways!

Of the two backup jobs that hung with 'Building VMs list' message one went through successfully at the next+1 scheduled run. The second job continued to fail. I created a new job to replace the failed one and it goes through without errors.

I should add that this is the first time I have encountered this problem despite having migrated dozens of VMs and individual disks in the past. Though none of this activity is ever undertaken while Veeam is running a backup.

Thanks again.

[Matthias]

Hi

same problem here.
After SvMotion of two VMs, both Backup jobs are hanging.
Veeam Server reboot does not resolve the issue.
The job hangs ands the process "VeeamAgent.exe" takes 100% of CPU.
For me it seems like a bug.
I guess this happens if you do a Storage vMotion of a VM which is regulary backed up and has changed block tracking enabled.

Is there anyone at Veeam who can say something about this issue?

Thanks
Matthias

[foggy]

Matthias, please consider a couple of advises given by Rich in his post above and contact our support team for further investigation. Thanks.