Libvirt lxc kills guest containers on first restart

Bug #1655534 reported by Christian Ehrhardt 
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
libvirt (Debian)
Fix Released
Unknown
libvirt (Ubuntu)
Triaged
Low
Unassigned

Bug Description

On the Tail end of working on https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=848317 I found that we have an issue with libvirt-lxc guests.

Effectively on the first restart of libvirt we loose all the guests.
I reported upstream and the soon emerging http://www.spinics.net/linux/fedora/libvir/msg141665.html so much sounded like the solution, but then if applied does not fix it.

Yet it is so close that the goup setup might be slightly different for us, but very likely this is the root-cause.

I'm opening this bug to track this, since we need a deeper check to be done why the upstream solution doesn't work for us.

While being a bad bug in terms of really breaking stuff, prio is only medium/low as to manage containers lxd is so much better and more recommended.

Changed in libvirt (Ubuntu):
status: New → Triaged
importance: Undecided → Low
Changed in libvirt (Debian):
status: Unknown → Fix Committed
Changed in libvirt (Debian):
status: Fix Committed → Fix Released
Revision history for this message
Christian Ehrhardt  (paelzer) wrote :

FYI - upstream fix 44f79a0b
Author: Daniel P. Berrange <email address hidden>
lxc: ensure libvirt_lxc and qemu-nbd move into systemd machine slice

Should be fixed completely when merging libvirt >=3.0

It seems no one bothered for quite a while - so I keep low prio.
The tests that trigger that have a band-aid for now that we can retest and drop when merging the newer libvirt (probably next cycle).

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.