Maybe I wasn't supposed to do this, but I had a working Win10 VM that was copied from a standalaone desktop hard drive.  Everything was working well, including passing through an AMD video card.  Decided it needed more horsepower, so I added some cores and RAM in its settings, and now it refuses to start.  It usually hangs on the windows logo, sometimes attempts to go into repair and reports that it cannot be repaired.   SOmetimes it freezes on the logo.  

 

Is it not advisable to change these settings on an established VM?

Try assigning just a single core and see what happens

Sent from my NSA monitored device

It loops on Preparing Automatic Repair

I have narrowed this down significantly.  I have been following the steps listed here.  Although I have not given it much time yet, it seems pretty clear that the new VM boots and runs fine if the primary vdisk bus is left set to IDE, but not if it is changed to VirtIO per the instructions.  It will go into a boot loop after blue screening with the stop code "Inaccessible Boot Device".  If I force stop and switch it back to IDE, Win10 comes back up as if nothing happened.  I did realize that I was using outdated VirtIO drivers, which were updated via Windows in the VM itself.  Not sure if that matters, but the boot issue persisted.

 

Is VirtIO a necessary step to enhance performance? Seems to run okay without it, and I havent immediately noticed any impoact to other stuff that UNRAID handles.  Is IDE a power hog? 

@btrcp2000 Add a second dummy vdisk let's say only 1G as Virtio. Boot your Windows and install the driver for that 1GB Harddrive. This way you can now remove the dummy file and switch your main vdisk to virtio and Windows should use the newly installed driver.

that did it, thanks!