Uploaded image for project: 'Red Hat OpenStack Services on OpenShift'
  1. Red Hat OpenStack Services on OpenShift
  2. OSPRH-17285

Database management jobs configured but not running

XMLWordPrintable

    • Hide

      None

      Show
      None
    • False
    • ?
    • rhos-workloads-compute
    • None
    • Important

      To Reproduce Steps to reproduce the behavior:

      1. Deploy an overcloud, 16.2 or 17.1
      2. Nova has the following database management jobs have been created by default in nova_api_cron:
      [root@brq-ctrl01 ~]# podman exec nova_api_cron tail -6 /var/spool/cron/nova
      # Puppet Name: nova-manage db archive_deleted_rows
      PATH=/bin:/usr/bin:/usr/sbin SHELL=/bin/sh
      1 0 * * * sleep `expr ${RANDOM} \% 3600`; nova-manage db archive_deleted_rows  --max_rows 1000 --before `date --date='today - 90 days' +\%F` --until-complete --all-cells >>/var/log/nova/nova-rowsflush.log 2>&1
      # Puppet Name: nova-manage db purge
      PATH=/bin:/usr/bin:/usr/sbin SHELL=/bin/sh
      0 5 * * * sleep `expr ${RANDOM} \% 3600`; nova-manage db purge --before `date --date='today - 14 days' +\%D`                      --all-cells >>/var/log/nova/nova-rowspurge.log 2>&1
      [root@brq-ctrl01 ~]#  
      1. Cinder has the following database management jobs have been created by default in cinder_api_cron:
      [root@brq-ctrl01 ~]# podman exec cinder_api_cron tail -1 /var/spool/cron/cinder
      1 0 * * * sleep `expr ${RANDOM} \% 3600`; cinder-manage db purge 30 >>/var/log/cinder/cinder-rowsflush.log 2>&1
      [root@brq-ctrl01 ~]#  
      1. None of the output files have anything in their logs, nor have them been rotated:
      [root@brq-ctrl01 ~]# ls -ltr /var/log/containers/{nova/nova-manage.log*,nova/nova-rows*,cinder/cinder-rowsflush.log*}
      -rw-r--r--. 1 42407 42407 0 Jul  8  2023 /var/log/containers/cinder/cinder-rowsflush.log
      -rw-r--r--. 1 42436 42436 0 Jul  8  2023 /var/log/containers/nova/nova-rowspurge.log
      -rw-r--r--. 1 42436 42436 0 May  2  2024 /var/log/containers/nova/nova-manage.log
      -rw-r--r--. 1 42436 42436 0 May  2  2024 /var/log/containers/nova/nova-rowsflush.log
      [root@brq-ctrl01 ~]# 

      Expected behavior

      • Jobs run and log output 
        •  

      Bug impact

      • This issue was found while troubleshooting errors in a customer upgrade, and is (allegedly) cause of database performance issues.

      Known workaround

      • Manual run of the jobs complete successfully

      Additional context

      • See customer case attached for extended details

              ksambor@redhat.com Kamil Sambor
              rhn-support-enothen Eric Nothen
              rhos-workloads-compute
              Votes:
              0 Vote for this issue
              Watchers:
              6 Start watching this issue

                Created:
                Updated: