Issue #2954
closed
Ensure that queued tasks are not lost by enabling task_reject_on_worker_lost for Celery 4
Description
In Celery 3, the resource_manager queue loses a currently running _queue_reserved_task if the resource manager is restarted with sudo systemctl restart pulp_resource_manager.
The task is lost from the queue but still has an incorrect TaskStatus record showing as waiting which will never run.
Note that if you sudo pkill -9 -f resource_manager and the sudo systemctl start pulp_resource_manager it does not lose the task.
sudo systemctl stop pulp_workers
pulp-admin rpm repo sync run --repo-id zoo
qpid-stat -q <<-- observe that the queue depth of the resource_manager queue is 1
sudo systemctl restart pulp_resource_manager
qpid-stat -q <<-- observe that the queue depth of the resource_manager queue is 0
pulp-admin tasks list -s waiting <<-- observe that the task which is gone is listed as 'waiting', but it will never run because it is gone
We need to make sure that this doesn't happen in Celery 4. There's a config task that should prevent this:
http://docs.celeryproject.org/en/latest/userguide/configuration.html#task-reject-on-worker-lost
Also, we to apply this fix for Pulp 2 AND 3.
Related issues
Updated by daviddavis over 7 years ago
- Subject changed from Queued tasks will be lost if the resource manager is restarted via systemctl for Celery 4 to Ensure that queued tasks are not lost by enabling task_reject_on_worker_lost for Celery 4
Updated by daviddavis over 7 years ago
It looks like tasks are currently being requeued for Celery 4 even if task_reject_on_worker_lost is not set. They shouldn't be and we need to investigate and potentially open an issue against Celery.
Updated by daviddavis over 7 years ago
- Related to Issue #2734: task group never quite finishes if pulp is restarted in the middle of task run added
Updated by daviddavis over 7 years ago
- Related to deleted (Issue #2734: task group never quite finishes if pulp is restarted in the middle of task run)
Updated by daviddavis over 7 years ago
- Related to Issue #2861: Queued tasks will be lost if the resource manager is restarted via systemctl added
Updated by daviddavis over 7 years ago
- Status changed from NEW to ASSIGNED
- Assignee set to daviddavis
Updated by ttereshc over 7 years ago
- Sprint/Milestone set to 42
- Triaged changed from No to Yes
Updated by daviddavis over 7 years ago
So I was able to reproduce the behavior in standalone celery where messages are persisted on warm shutdown even if task_reject_on_worker_lost is not set. It turns out that if you run pkill -f celery instead of kill $CHILD_PROCESS_ID, the message gets persisted.
This is why when shutting down pulp_resource_manager via systemctl, we're seeing messages getting persisted. It's killing (or doing a warm shutdown) on both processes. I have no idea why this is. I can open an upstream celery issue but this behavior sounds pretty much the same as some existing bugs [1][2].
Also, the message persisting is not a problem for us though. We're concerned about messages being lost and if a message gets persisted, it simply runs the next time pulp starts up. We're not concerned here about double execution either since pulp_workers are not running.
[1] https://github.com/celery/celery/issues/3802
[2] https://github.com/celery/celery/issues/3796
Updated by daviddavis over 7 years ago
So basically we just need to enable task_reject_on_worker_lost and we're good to go.
Updated by daviddavis over 7 years ago
- Related to Issue #2958: Ensure that queued tasks are not lost by enabling task_reject_on_worker_lost added
Updated by daviddavis over 7 years ago
Opened separate issue for Pulp 3:
Updated by daviddavis over 7 years ago
- Status changed from ASSIGNED to POST
Updated by daviddavis over 7 years ago
I would probably recommend using the following workflow for testing as it's a bit more precise in that it only kill the child worker process. Using sudo systemctl restart pulp_resource_manager will kill both the child and the parent which will potentially leave the message in the queue and thus would be a false positive.
sudo systemctl stop pulp_workers # may need to wait 30 seconds for this to die
pulp-admin rpm repo sync run --repo-id zoo --bg
qpid-stat -q # observe that the queue depth of the resource_manager queue is 1
ps auxf | grep resource_manager # grab the child process id (e.g. 12345)
sudo kill 12345
qpid-stat -q # observe that the queue depth of the resource_manager queue is still 1
sudo systemctl restart pulp_resource_manager
sudo systemctl start pulp_workers # may need to wait 30 seconds for this to start and pick up task
pulp-admin tasks list -s waiting # should be empty
Added by daviddavis over 7 years ago
Updated by daviddavis over 7 years ago
- Status changed from POST to MODIFIED
Applied in changeset pulp|af4f688aec41908855c02d4eb709c75e116dc5f3.
Updated by pcreech about 7 years ago
- Status changed from MODIFIED to CLOSED - CURRENTRELEASE
.
Turn on task_reject_on_worker_lost to prevent lost tasks
Turn on task_reject_on_worker_lost (aka CELERY_REJECT_ON_WORKER_LOST) to prevent the loss of tasks when a worker dies. This option is only available in Celery 4+.
fixes #2954 https://pulp.plan.io/issues/2954