Red Hat Bugzilla – Bug 1276380
NPE while running a VM with passthrough devices
Last modified: 2016-02-10 14:24:07 EST
Created attachment 1087581 [details]
I was playing with UI and found an issue when removing assigned devices. NPE in the log and weird behavior
Time: 2015-10-24 09:05:02,871
I had a VM with several devices assigned and only one host (not properly configured for passthrough) and tried to Run it
Time: 2015-10-24 09:06:05,921 and around:
Then I tried to remove the devices (not all at once, in 3 steps) and during the removal of the last one I received an empty result, device disappeared from the UI but apparently remained in backend - after switching tabs the device was still there
seems it left the VM in some inconsistent state as it can't be run ever again even when all devices are removed and engine is restarted
The issue of failing to run the VM has already been fixed in 3.6. Marking as duplicate.
The '' in devices removed may appear in cases when you issue the remove commands too quickly and the device has already been removed as a result of previous bulk command. In that case the second operation is NO-OP thus resulting in "empty list of devices has been removed" message.
*** This bug has been marked as a duplicate of bug 1256238 ***