yahoo-eng-team team mailing list archive

[OpenStack Infra <1784874@xxxxxxxxxxxxxxxxxx>]

Reviewed:  https://review.openstack.org/587922
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=99db9faae5c1e84e5aa8586c32ffcca7a61ae276
Submitter: Zuul
Branch:    master

commit 99db9faae5c1e84e5aa8586c32ffcca7a61ae276
Author: Dan Smith <dansmith@xxxxxxxxxx>
Date:   Wed Aug 1 10:16:41 2018 -0700

    Hook resource_tracker to remove stale node information
    
    When we remove a node from a host due to rebalance or decommission,
    we need to tell the resource tracker so it stops keeping track of
    the details for that node. This adds a hook for doing that and
    calls it from the place in compute manager where we purge the
    record from the database.
    
    Change-Id: Ie6b6bb2a9e8a4ad33675fccb3827e8197fa16398
    Closes-Bug: #1784874


** Changed in: nova
       Status: In Progress => Fix Released

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Compute (nova).
https://bugs.launchpad.net/bugs/1784874

Title:
  ResourceTracker doesn't clean up compute_nodes or stats entries

Status in OpenStack Compute (nova):
  Fix Released

Bug description:
  This was noted in review:

  https://review.openstack.org/#/c/587636/4/nova/compute/resource_tracker.py@141

  That the ResourceTracker.compute_nodes and ResourceTracker.stats (and
  old_resources) entries only grow and are never cleaned up as we
  rebalance nodes or nodes are deleted, which means it just takes up
  memory over time.

  When we cleanup compute nodes here:

  https://github.com/openstack/nova/blob/47ef500f4492c731ebfa33a12822ef6b5db4e7e2/nova/compute/manager.py#L7759

  We should probably call a cleanup hook into the ResourceTracker to
  cleanup those entries as well.

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1784874/+subscriptions