Comment 4 for bug 1658629

Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to nova (master)

Reviewed: https://review.openstack.org/424305
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=b2a4fcf84633413c322afb4f2c8066358b943d6b
Submitter: Jenkins
Branch: master

commit b2a4fcf84633413c322afb4f2c8066358b943d6b
Author: Chris Dent <email address hidden>
Date: Mon Jan 23 19:34:44 2017 +0000

    Avoid redundant call to update_resource_stats from RT

    When the resource_tracker calls _update, when the new compute node
    and the old compute node only differ by updated_at,
    update_resource_stats was still being called. Manual testing shows
    that updated_at is generally different because there's something
    else that is also leading to a compute_node.save() approximately
    every sixty seconds. So while this fix removes one redundant
    save, it doesn't get all of them, thus the "partial" below.

    A unit test is added to exercise different updated_at values.

    Change-Id: If688ae4d92ecdea83479f37de8856b668b8bc7a6
    Partial-Bug: #1658629