Case #04282072 - Virtual RDM is treated as physical RDM by Veeam

[ASG]

Hi,

After a lot of searching (the first effect we've noticed was that virtualLab timed out when a specific VM is present) it was found out that Veeam treats our 2 virtual-dependent RDMs for this VM as physical and on Virtuallab/FLR etc... first tries to restore the about 6,5TB. That's an error - as the disks are clearly attached as virtual-dependent. Veeams backs them up without any warning or what-so-ever.

I've tried searching but couldn't find something in the same direction. Has anybody had the same problem? Any chance to fix this without throwing the backups away?

Edit: We have several other VMs with virtual-dependent that work as the should without problems

Edit2: When creating a new backup of this VM the 2 RDMs get backed up as virtual-dependent and can be mounted without a problem.

[Andreas Neufert]

This case is under investigation by Veeam Support. Please check with them directly.

[ASG]

So - it's a bit complicated.

Looooong story short:
When you change something to RDMs use different SCSI IDs or make an active full. Turns out B&R took our vrdm for the prdm it has seen at the first backup (configuration error that veeam throw an email about that was corrected on the same day) - but since vmware attached the vrdms with the same SCSI IDs it "was the same disk" to B&R - ignoring what the discriptor file said. Switching the SCSI IDs fixed it for new backups (after reading several TB data).

[Andreas Neufert]

Thanks for following up and adding the explanation.