incorrect changed block tracking configuration

[Oletho]

I have a strange CBT problem.

Running a mixed ESX/ESXi 4.1 cluster with Veeam 5.0 in VA mode.

A couple of weeks ago each day I had 1-2 VMs failing CBT with this error:

Verifying changed block tracking...
Disk "Hard disk 2" has incorrect changed block tracking configuration.
Disk "Hard disk 3" has incorrect changed block tracking configuration.

As it was different VMs each time, and Veeam recently was upgraded to 5.0 I just went on and repaired CBT. Soon the errors stopped.

But last night I had a dozen VMs giving me the same error. The VMs were from different hosts and Veeam jobs.

I don't think it is a problem with storage, as none of them were failing with disk 1 but on all other disks. I suspect a bug or misconfiguration somewhere, but cannot find anything wrong.

Ole Thomsen

[Gostev]

Yes, it looks like something is braking your source VMs configuration - Veeam Backup simply queries it before backup, and notifies you that it is incorrect.

[Oletho]

Tonight I had the error on a VM that has been turned off for months. I cannot imagine that the tracking file has changed since the previous backup

Ole Thomsen

[Gostev]

VM power state definitely does not matter. Changed block tracking configuration which we are validating before backup starts are stored in VM configuration. They can be changed by anything without powering VM on. They may also be affected by performing some reconfiguration operations on the VM without powering it on (for example, adding new disk). Thanks!

[Oletho]

This is really confusing.

I now have a job where 11 VMs has the Incorrect... message in the report, and Veeam states that it runs these without CBT.

But this job has the same runtime as it had with no CBT warning, around 11 hours?

Any suggestions how I troubleshoot this - should I open a support ticket with Veeam? With VMware?

Ole Thomsen

[Gostev]

Hi Ole, you can always start with Veeam support, they will tell you if you need to contact VMware after investigating specific issue. Thanks!

[Oletho]

I just opened a ticket with support.

Would like to add that after I stopped repairing using advanced settings, the error goes away for different VMs. And eventually comes back.

Ole Thomsen

[Gostev]

Sounds good Ole, I have also asked dev lead to investigate possibility of fully automatic "CBT fixing" to the next release. Thanks!

[mplep]

Hi. If I receive this error, does Veeam still backup the disk, but in Full mode rather than Incremental?

Thanks.

[Vitaliy S.]

Your disk will still be backed up by a legacy change block tracking engine, but it will take time comparable to a Full run (a little faster).