Worked locally - less sure on prod!
To Pause
drush @wmff cvapi Setting.create queue_paused=1
To unpause
drush @wmff cvapi Setting.create queue_paused=0
Description
Description
Related Objects
Related Objects
Event Timeline
Comment Actions
I might bring this into our current sprint as I have wound up having to deal with coworker stuff - moving to triage for now
Comment Actions
@Dwisehaupt I confirmed that this DOES work on prod - ie after running drush @wmff cvapi Setting.create queue_paused=1 then no further queue items were claimed once the claimed item is completed - do we need to update our outage practices?
It's possible to check with this queue
select queue_name, max(release_time), count(*), max(run_count), status FROM civicrm_queue_item i LEFT JOIN civicrm_q
ueue q ON q.name = i.queue_name WHERE release_time IS NOT NULL GROUP BY queue_name;
+---------------------+---------------------+----------+----------------+---------+
| queue_name | max(release_time) | count(*) | max(run_count) | status |
+---------------------+---------------------+----------+----------------+---------+
| omni-snooze | 2024-02-20 03:07:49 | 1 | 16 | active |
| omni-snooze/damaged | 2024-02-19 12:07:33 | 15 | 52 | aborted |
| user_job_1114 | 2024-02-20 03:22:59 | 1 | 1 | active |
+---------------------+---------------------+----------+----------------+---------+
& I updated the docs
https://wikitech.wikimedia.org/wiki/Fundraising/Internal-facing/CiviCRM#Queue_processing_(coworker)