We're running a 2019 Hyper-V cluster. During a patch operation the host was not drained of all roles before paused and rebooted. This has resulted in a virtual machine that can't be migrated to the specific host again. The mentioned host gets an error at boot (when VMMS starts) because it tries to find a missing virtual machine. The virtual machine resides on another host in the cluster.
We've tried stopping VMMS and removing files in "%Programdata%\Microsoft\Windows\Hyper-V\Virtual Machines Cache" but the files keep re-appearing once VMMS starts up again.
Where else should we look for this ghost entry?