Comment 19 for bug 1714248

Reviewed: https://review.openstack.org/508555
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=e3c5e22d1fde7ca916a8cc364f335fba8a3a798f
Submitter: Zuul
Branch: master

commit e3c5e22d1fde7ca916a8cc364f335fba8a3a798f
Author: John Garbutt <email address hidden>
Date: Fri Sep 29 15:48:54 2017 +0100

    Re-use existing ComputeNode on ironic rebalance

    When a nova-compute service dies that is one of several ironic based
    nova-compute services running, a node rebalance occurs to ensure there
    is still an active nova-compute service dealing with requests for the
    given instance that is running.

    Today, when this occurs, we create a new ComputeNode entry. This change
    alters that logic to detect the case of the ironic node rebalance and in
    that case we re-use the existing ComputeNode entry, simply updating the
    host field to match the new host it has been rebalanced onto.

    Previously we hit problems with placement when we get a new
    ComputeNode.uuid for the same ironic_node.uuid. This reusing of the
    existing entry keeps the ComputeNode.uuid the same when the rebalance of
    the ComputeNode occurs.

    Without keeping the same ComputeNode.uuid placement errors out with a 409
    because we attempt to create a ResourceProvider that has the same name
    as an existing ResourceProvdier. Had that worked, we would have noticed
    the race that occurs after we create the ResourceProvider but before we
    add back the existing allocations for existing instances. Keeping the
    ComputeNode.uuid the same means we simply look up the existing
    ResourceProvider in placement, avoiding all this pain and tears.

    Closes-Bug: #1714248
    Co-Authored-By: Dmitry Tantsur <email address hidden>
    Change-Id: I4253cffca3dbf558c875eed7e77711a31e9e3406