Uploaded image for project: 'Red Hat OpenStack Services on OpenShift'
  1. Red Hat OpenStack Services on OpenShift
  2. OSPRH-27488

nova-api is_up gets incorrect service status for nova-compute - "Seems service nova-compute on host nova-compute01 is down"

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Unresolved
    • Icon: Undefined Undefined
    • None
    • None
    • openstack-nova
    • None
    • Low

      rhoso18 environment.

      From nova-api log:

      2026-03-10 16:26:31.244 11 DEBUG nova.servicegroup.drivers.db [None req-e29ba2da-ea56-4faf-b3b8-515ad35660c2 3e718b86717142ccbb82f5a96c57db13 faac9d4b644844408ff6cc7178e54a3f - - default default] Seems service nova-compute on host nova-compute01 is down. Last heartbeat was 2026-02-19 20:47:17. Elapsed time is 1625954.244931 is_up /usr/lib/python3.9/site-packages/nova/servicegroup/drivers/db.py:76

      This issue occurred when trying to do a volume snapshot of an in-use volume.

      nova-compute services generally worked fine; a VM was created with volume attachments just prior to this failure.

      However, nova-compute service is up. The service status seems to have come from an old soft-deleted service record in the db.

      sh-5.1$ openstack compute service list |grep nova-compute

      6357ba69-8961-46a8-b34f-fb88f80afe63 nova-compute nova-compute01 nova enabled up 2026-03-09T16:14:24.000000
      5e13cc19-b6d9-45ff-b6a8-3df0bc595282 nova-compute nova-compute00 nova enabled up 2026-03-09T16:14:24.000000

      select updated_at, deleted_at, deleted, host, `binary`, last_seen_up from services where host = 'nova-compute01';
      --------------------------------------------------------------------------------------------------------------+

      updated_at deleted_at deleted host binary last_seen_up

      --------------------------------------------------------------------------------------------------------------+

      2026-02-19 20:47:17 2026-02-20 19:13:06 3 nova-compute01 nova-compute 2026-02-19 20:47:17
      2026-02-23 17:25:23 2026-02-23 20:18:10 4 nova-compute01 nova-compute 2026-02-23 17:25:23
      2026-02-24 17:17:37 2026-02-24 17:17:41 6 nova-compute01 nova-compute 2026-02-24 17:17:37
      2026-02-24 17:25:58 2026-02-24 17:26:06 8 nova-compute01 nova-compute 2026-02-24 17:25:58
      2026-02-24 17:30:20 2026-02-24 17:30:23 10 nova-compute01 nova-compute 2026-02-24 17:30:20
      NULL 2026-02-24 19:38:47 12 nova-compute01 nova-compute NULL
      2026-03-06 16:56:12 2026-03-06 19:37:08 14 nova-compute01 nova-compute 2026-03-06 16:56:12
      2026-03-10 17:00:48 NULL 0 nova-compute01 nova-compute 2026-03-10 17:00:48

      --------------------------------------------------------------------------------------------------------------+

      When the 'deleted' service records were purged the above operation completed successfully.

      delete from services where deleted > 0;

      Seems this is_up operations should exclude deleted services.

              Unassigned Unassigned
              mflusche@redhat.com Mathew Flusche
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

                Created:
                Updated: