Comment 12 for bug 1833096

Reviewed: https://review.opendev.org/677706
Committed: https://git.openstack.org/cgit/starlingx/nfv/commit/?id=4ea74a99c9148ee5c813a3bec086103c463634e5
Submitter: Zuul
Branch: r/stx.2.0

commit 4ea74a99c9148ee5c813a3bec086103c463634e5
Author: Bart Wensley <email address hidden>
Date: Fri Aug 16 13:48:09 2019 -0500

    Make VIM tolerant of compute service failures

    When the VIM detects the nova compute service is down on a
    worker host, it attempts to migrate instances off that host
    (by "disabling" the host). However, this isn't possible if
    the compute service is down. The VIM then fails the instances,
    which will eventually result in their evacuation (if the host
    goes offline) or a reboot of the instance (if the compute
    service recovers).

    In the containers world, when the libvirt pod is restarted
    (e.g. when stx-openstack application is re-applied), nova
    reports that the compute service is down (for a short period
    of time), which causes the undesirable behaviour described
    above. The VIM is being updated to not disable the host in
    this case and instead just raise an alarm to indicate that
    the compute service has failed.

    Change-Id: I186d8d76bbcd87405bafec47deb92ec24580640e
    Closes-Bug: 1833096
    Signed-off-by: Bart Wensley <email address hidden>
    (cherry picked from commit a9004988dc37bdd9caefdbcf911472b38c4db5ac)