Hyper-V cluster: SAIPH has PhysHostId = Zero after NIC replacement

Did you try editing the Hyper-V host/cluster in the Veeam console?  Edit and click through the wizard to see if it re-registers it?  If not you might need support to help with this one.

Thanks Chris for taking the time to reply.

Yes, we tried the Edit Hyper-V Server wizard. This is actually where we found the connection problem.

When we proceed with the existing SAIPH entry, the wizard fails at the Credentials step with:

“Failed to connect to host saiph”
“Server ‘saiph’ does not exist or access denied.”

We also verified that the credentials themselves are valid, without exposing any credentials or domain details:

• We used the same account configured in Veeam.

• On SAIPH, we authenticated with that account using runas.

• The password was accepted and a new CMD session was opened under that account.

• whoami confirmed the expected administrator account.

So the credentials appear to be valid, but Veeam still cannot establish the Hyper-V connection to SAIPH through the wizard.

We have attached a screenshot of the Veeam wizard showing the failure.

Thanks again for pointing us in this direction. Any further thoughts on what could prevent the Hyper-V connection/registration despite valid credentials would be greatly appreciated.

Off hand I cannot think of anything sorry.  We don't use Hyper-V other than a few clusters but they are simple setups.  Support should be able to get it sorted though.

One other thing to try is the username as UPN versus domain\username.  Use user@domain.com maybe this will work but not guaranteed.

Hi ​@tolisfil , 

I would still try with a Support Case even if it’s Community Edition, as the situation is a bit challenging.

The issue is purely in the configuration database and you’re spot-on with the physical_host_id issue; however, correcting the physical_host_id is not the full solution here, as whatever caused it to get the 0’d physical_host_id may trigger again, so important part is to figure out why that happened.

From experience, usually issues reading information from the host itself will cause our resync task to 0-out a physical host id.

If you want to poke on your own, you might try:

1. Do a manual rescan of the cluster
2. Once complete, check out %ProgramData%/Veeam/Backup/Utils/VolumeHostDiscovery.log (there may be many of them, look at the most recent one)
3. Find the latest session (search for new log and jump to the last one), and start reading

Something is messing up the detection process, so it’s a matter of figuring out what. Cannot really give a guess as to what it is, and really Support is best for this, but you can at least try your hand at some log diving.