Hi Alex,

1a. handle_virt_lifecycle_events might just help nova to get less
confused about it - on the actual reboot it should make no difference -
unless nova is no e.g. no more issuing double-restarts

1b. Given the former "incarnations" of this issue I could not yet
recommend a way to get better debug - the problem is that when the
problem occurs things are already broken, and to catch some change
before where you don't know yet if it is the bad one is much harder

1c. You can if you add UCA as I linked above just update selected packages - 
some of them might have dependencies but apt will handle that for you.
So you could do the add-apt-repository, then you check "apt list --upgradable" 
and likely only select "apt install qemu-system-x86 libvirt-bin"
That will pull in qemu/libvirt and those dependencies that are absolutely 
necessary, but not all of openstack form that archive.
OTOH - the older openstack you have atm might have issues to handle the newer 
libvirt/qemu. Worth a try if you can so so in your environment.

2. thanks, I had all sorts of reboot loops not triggering it. Unless
nova issues soemthing else to reboot that should be ok.

3. thanks - in that regard network I/O is I/O as well - so webserver
with low disk I/O is what I will remember from now on.

Looking forward to hear if a newer virt stack gets you fixed ...

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1721167

Title:
  Guest stopped after "Virtqueue size exceeded error"

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1721167/+subscriptions

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to