Symptoms

I've encountered this error after upgrading a RHV 4.3 environment to 4.4 and since this release already permits updating both the cluster and the data center to 4.5 (I'd have expected 4.4) I tried and failed a couple of times because upgrading the cluster level to 4.4 always failed because of a couple of VMs.

Looking at the messages about each VM in the UI didn't really help much either:

Cannot update compatibility version of Vm/Template: [<name-of-the-vm], Message: [No Message]

Identifying the reason

Search for this error message didn't reveal much useful at all. So i turned to engine.log on the RHV manager which has a tendency of being very verbose, thus not really helpful.

So I let run tail -f /var/log/ovirt-engine/engine.log | tee /tmp/engine-extract.log and attempted another upgrade.

Searching for the name of the listed VMs.only gave me:

2021-04-xx xx:xx:xx,xxxZ ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (default task-2) [3e664583] EVENT_ID: CLUSTER_CANNOT_UPDATE_VM_COMPATIBILITY_VERSION(12,005), Cannot update compatibility version of Vm/Template: [beaufix], Message: [No Message]

Allas No Message isn't really that much more helpful either.

However bat at least limiting the amount of logs to dig through I finally stumbled upon the following kind of messages:

2021-04-xx xx:xx:xx,xxxZ WARN  [org.ovirt.engine.core.bll.UpdateVmCommand] (default task-2) [60ce7f39] Validation of action 'UpdateVm' failed for user <user>@<realm>-authz. Reasons: VAR__ACTION__UPDATE,VAR__TYPE__VM,ERROR_CANNOT_FIND_ISO_IMAGE_PATH

The amount of occurences of these messages matched with the number of VMs that were listed in the initial error message.

At that point it was clear that the affected VMs had some lingering ISO mounts configuredfor ISO images not present anymore on the ISO storage domain.

Checking if a ISO was still mounted revealed nothing though.

Fixing

In the end I had to edit each VM settings and detach the virtual CD/DVD drive (which is possible while the VM is online), simply ejecting wasn't enough.

Afterwards bumping the cluster, then the datacenter level to 4.5 worked fine.

However I whished the log messages had been "a tad" more helpful in the first place.

TL;DR: If you get a not very helpful message in the oVirt or RHV UI, don't hesitate to dig through the engine Logs of the manager.

Unfortunately they are both rather verbose and not that easy to dig through.at the same time.