+3
Hello
Recently our Windows VSA proxies started to bluescreen during backup when we are using HotAdd as transport Mode. Switching to NBD works fine, but is a lot slower.
Happens in 2 different ESX cluster running both vSphere 7.3 and 8
The server BSOD with message
HYPERVISOR_ERROR (20001)
The hypervisor has encountered a fatal error.
Arguments:
Arg1: 0000000000000013
Arg2: 0000000000000000
Arg3: 0000000029b92701
Arg4: ffffe800005492c0
Anyone else experienced recent issues with HotAdd backups?
+19
Any idea if automount was disabled on the VSA?
Diskpart
automount
Automount disable
automount scrub
+11
Hello
Can you please confirm if automount is disabled?
diskpart> automount
diskpart> automount disable
diskpart> automount scrub
Also, please confirm if the SCSI ports are in order:
https://documentation.commvault.com/2023e/expert/32048_hotadd_transport_for_vmware.html
To end with, please check if all the drivers are upto date on the proxy server.
Best,
Rajiv Singal
Hi
Finally try : https://documentation.commvault.com/2023e/expert/32048_hotadd_transport_for_vmware.html#testing-hotadd
You can use the following process to test HotAdd attachments for a VM:
Create a snapshot on the VM that you planning to back up.
Attach the base disk from that VM to the VSA proxy as an independent or non-persistent disk.
If the attach is successful, the VSA proxy will be able to perform a HotAdd operation.
Detach the disk from the VSA proxy.
Remove the snapshot for the VM.
If you can reproduce the issue outside Commvault then you can log a case with Vmware.
Best Regards,
Sebastien
+3
Automount is disabled. Well, it wasn’t from the beginning but enabling it did not help
This used to work fine, but started to appear now in October. I guess some Windows update is causing this, or maybe ESX, but I get the same error on 2 different clusters with different versions of ESX
I will open a support ticket for this, and also try to HotAdd manually
Thanks
+19
Any other update that has been done like AV ?
+3
Hello
Not really easy to deploy a Windows server without required updates. One thing I have noticed is that it looks like it was specific VM’s that caused the crash, However suddenly I was able to do a HotAdd of one of them, so still no clear picture of what is causing this.
I think I will look at using a Linux based access node instead, initial tests are promising
+19
Ok. Not sure if you looked into using FREL, but we use it for a long time already and the nice thing is that you can deploy them directly from Command Center.
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
