Cluster Change Block Tracking

[hafedh]

Hello everyone,

I have installed VEEAM B&R V9 and i have a Warning when i try to Backup Hyper-v Virtual Machine (This Virtual machine is in a Cluster HyperV 2008 R2);
Warning: Failed to flush change tracking data. Using full scan incremental run.
I don't have the possibility to backup this virtual machine with incremental Backup, just Full.

If a try to hyperv virtual machine (Standalone Hyperv - Not clustered), it's fine and i don't have any problem.

Can anyone help me please, thanks.

Best regards,
Hafedh

[Vitaliy S.]

Please take a look at this KB article and existing topic for more info on how to troubleshoot it. Thanks!

[hafedh]

Thanks a lot for your reply,

I will check all of thsese solutions and I would come back to you to inform you of the result.

Thanks.

[hafedh]

Hi,
I have a question please: One of the cluster-nodes is turned off, this can be the problem ?

Thanks;

[Vitaliy S.]

Yes, CBT data is stored on each node of the cluster, and if it is turned off, then VM backup job will do a full image scan.

[marieelieg]

Hello everyone. I am quite new to the forum so please excuse any wrongdoing on my part.

I have a hyper-v cluster and I created a backup containing some VMs. Some of those VMs didn't have hyper-v integration services up-to-date, but the backup was successful. I then updated the integration services on the VMs and rerun the backup (now it's incremental). Some of the VMs gave a warning :"Failed to create change tracking time stamp for virtual disk file" but the backup was completed. My question is when upgrading the integration services, does it reset or affect the CBT information in any way and that's why the backup gave me a warning? And if so do I have to run the backup job again as a full backup or does an incremental backup be sufficient?

I really appreciate your feedback as soon as possible and thank you in advance:D

[Mike Resseler]

Hi,

It depends actually. I have seen CBT resets for various reasons. If you do a failover to another host, if there is a sudden shutdown of the host (and then the host loses it's CBT information) or, as you state, when integration tools are updated and depending on what is in the update (so it is not every time actually) it can reset this also.

Windows server 2016 hyper-v will fix this but I'm afraid the older server solutions will always have this possibility.

You don't need to do a new full backup, as we will continue to do an incremental, however, we need to scan the VM to see what the changes are, and therefore, you will notice that the backup takes longer.

Hope it helps

Mike

[marieelieg]

Thank you Mike. Your help is much appreciated.

[kshotrod]

Veeam 8, Hyper V cluster, it does this on every backup that is ran. Each backup is there, but the boss wants the warnings gone from the logs, so i need to try and fix this. I ran the powershell command to reset the bit, that didnt work...

[foggy]

Hi Charlie, some hints are given above, please review this thread.

[wayne@kesmi.com]

My Veeam Detail:

Ver : 9.5.0.1038
Backup Type: Incremental

Cluster with 2 Hosts , Per Host have 4 VM.


It happen when im shut down all server include veeam backup server because of power maintenance ,
after Im start back the all server including Veeam Backup it display 'Failed create change tracking time stamp for virtual disk file',


Im try to :

- Reboot Nodes, Hosts, VMs
- Stop VeeamFCT service and Start it again
- Running Reset-HvVmChangeTracking

but no lucky.Please help.Thank you.

[Shestakov]

Hello Wayne and welcome to the community!
That's a known Hyper-V cluster CBT issue. CBT data is stored on the hosts individually. Please review the thread for more details.

Have you tried to run backup job after all hosts are powered on?
Thanks!

[L33T]

Hi,

I have added a couple more hosts to my cluster but since then i get:

18/06/2018 16:20:32 :: Failed to create change tracking time stamp for virtual disk file C:\ClusterStorage\SERVER.VHDX

Any ideas?

[Shestakov]

Hi L33T,
That looks like cluster CBT case. Please review the topic.
Thanks!