Comment 3 for bug 1643911

I found this bug report while investigating gate issue with this query:

message: "kernel: traps: libvirtd" and message: "trap stack segment ip:" and message: "error:0 in libc-2.23.so" AND tags:"syslog"

I am not sure it's related to this issue, but so far what I know is that sometimes libvirtd crashes 'systemd[1]: libvirt-bin.service: Main process exited, code=killed, status=7/BUS' and is the restarted by systemd a few seconds latter.

The job status is not always FAILED because how Tempest works: in some tests, if deleting a server fails (VM status goes to ERROR), Tempest doesn't mark the test as failed: https://github.com/openstack/tempest/blob/fc267faa92c5355d9971b7dac3275ffea9646cd0/tempest/api/compute/base.py#L141