Instant Recovery/Centos7 : vm starts in recovery mode

[karim]

Good morning all,
I am using Veeam 10 and our virtual infrastructure is under vSphere / vSAN 6.7.

I have a problem with the instant recovery mode for some VMs. These virtual machines run under Centos 7, they have a system disk configured in LVM and a second data disk /dev/sdb without partitions.
When instantiating instant recovery, the VM starts in Recovery mode. If I restart the VM, it starts up in normal mode. I can't understand why. In recovery mode if I do CTRL + D it starts correctly and all the partitions are mounted. If I enter the root password for maintenance, I see no problem, the partitions are all present.
If I launch a new instant recovery session with the same restore point or another, the problem is still present.

I would like to point out that the problem does not appear on the production VM. I found this article which exactly describes my problem. https://access.redhat.com/solutions/3987281

a timeout story

I have the impression that this problem only appeared since our migration to vSAN.

I attach you the screenshots of the vm startup.

have you ever encountered this problem?
Thanks for your help.

[PetrM]

Hello,

Just a couple of questions: did you run a backup job with VMware Tools Quiescence enabled? Did you try to restore this VM from full restore point and does the same problem persist if you make Entire VM Restore to a different location?

Thanks!

[karim]

yes vmtools quiescence is enabled
i started the entire restoration, the data disk is 4TB in size, it will take some time.
I'll keep you posted tomorrow
thank you

[karim]

Hello
after 18h of restoration the vm starts without problem. would it not be a problem related to the bios or vm uuid?
I know that the option "regenerate uuid bios" is not functional in Powershell and will be corrected in the next version of the product (
my instant recovery procedure is executed by a powershell script.
thanks.

[PetrM]

Hello,

Ok, thanks for information. I cannot say what is the source of the issue without examination of debug logs but apparently it's somehow related to Instant Recovery itself. What is the result if you run Instant Recovery via UI to a different location, does the same issue occur? If not, it would make sense to dig deeper into PS script. Also, I recommend to open a support case and to share its ID over here as we're talking about technical issue.

Thanks!

[karim]

Hello
Indeed I also have the problem using the Veeam console and restoring to a different location (esx different from the vm of prod and a new name). i opened the Case #04855349
Thank you!