Backup without changed block tracking last result Warning

[Davd]

Hi,

I have a backup job with 15 VMs and where 2 of the VMs are powered off. These 2 VMs can be powered on and off at any time and therefore I still need to have them in the backup job. CBT doesn't work on powered off VMs and in Veeam 5 I could see in job statistics that both VMs where backed up without changed block tracking and the job status was Success. But in Veeam 6 those 2 VMs give job status Warning and I can see in details that the warning is casued by "Cannot use CBT: Soap fault. A specified parameter was not correct ..". If I power on the VMs and run a backup, CBT is used and status is Successful.

[Gostev]

Hi David, would you like this warning to be turned into the informational event in cases when the reason for CBT failure is the fact that VM is powered off?

[Davd]

Hi Gostev

That would be perfect... its always nicer to have all jobs status as Successful

btw... regarding version 6, the best way to summerize it: That's one small step for a man, one giant leap for virtual backups

There are so many great features that you guys added! Thanks for all you work!

[Gostev]

OK, we are planning to convert another warning into informational event in 6.0.1, so I will ask to change this one as well.
Thanks for your kind words, glad you are enjoying v6!

[jevao]

If my powered off VMs have only one disk, I do not see the CBT error below.
I only see following error on the second disk of a Powered Off VM.

"Cannot use CBT: Soap fault. A specified parameter was not correct. . deviceKeyDetail: '<InvalidArgumentFault xmlns="urn:internalvim25" xsi:type="InvalidArgument"><invalidProperty>deviceKey</invalidProperty></InvalidArgumentFault>', endpoint: '' error."

It seems for a powered off VM, if there are no changed blocks, the backup output for the first hard disk shows [CBT],
but 0 KB is read. On my VMs, the second disk actually reads some blocks, hence, even on powered off VMs,
it is only on the second hard disk where CBT fails. On the VM I checked,
the second disk says it is a page file when I boot it up, but not sure why it reads the blocks of a powered off VM?

[Vitaliy S.]

This behavior is going to be changed, so that only informational event will be triggered for powered off VMs, though I have no idea why only second disk causes this issue. This needs to be investigated.

[MattR]

I have the same cannot use CBT warning after migrating a host from one SAN datastore to another. The VM is powered on.

Code: Select all

1/17/2012 9:47:37 AM :: Cannot use CBT: Soap fault. Error caused by file /vmfs/volumes/<identifier>/Testing VM/Testing VM_1.vmdkDetail: '', endpoint: ''

I understand this will change to an informational note, but does the lack of CBT mean that the job will take longer to process? I am limited in space on the original SAN datastore for the host, which is why the vmdk files were moved. Is there a way to get back to CBT other than to migrate it back to the original datastore?

--

[Gostev]

Hi - actually, incremental backup will still work, but it will be much slower without CBT.
CBT fix process is described in the sticky Known Issues topic in the main forum.
Thanks.

[matthadley]

Ok. This morning when I started a replication job I started getting the following warning.

2/8/2012 9:39:22 AM :: Cannot use CBT: Soap fault. A specified parameter was not correct. . deviceKeyDetail: '<InvalidArgumentFault xmlns="urn:internalvim25" xsi:type="InvalidArgument"><invalidProperty>deviceKey</invalidProperty></InvalidArgumentFault>', endpoint: ''


I know of no reason that this should suddenly begin and apply to all four guests. That said, this is a replication of four guests, resident on an ESXi 4.1 host to an ESX 4 host. I can look at making the CBT fix but I should not have to. These are brand new replication jobs on a brand new installation (Monday AM) of Veeam 6 Enterprise. All four guests are up and running and have been.

Thoughts?

[matthadley]

Update - I followed the instructions to work the CBT fix and it appears to have worked. In the advanced settings, the individual entries, per hard disk, did not exist. I will consider this issue resolved from my perspective unless it re-appears on this four guests for reasons that I cannot fathom.