Delayed Jobs Issue after update 5.2.2

[KINGRULZ]

• Used Zammad version: 5.2.3
• Used Zammad installation type: (source, package, docker-compose, …) Package
• Operating system: Ubuntu 20.04
• Browser + version: Chrome: 106.0.5249.91

Expected behavior:

Process Jobs with issues

Actual behavior:

Zammad seems to stop processing jobs randomly, steps i have done
updated from 5.2.2 to 5.2.3
Reinstalled Ingest Attachment restart elastic,
rebuilt index
Remove Elasticsearch and reinstalled,

purged zammad and reinstalled

But background jobs will just randomly start back logging up to 20 - 100k of them
i simply restart zammad and they start working when the delayed jobs are above 15k i have to restart zammad-worker a few times to clear them, once cleared zammad will work perfectly for a while and then randomly they start building up again, i am now restarting zammad-worker several times a day,

the logs dont seem to show any issue and im not sure how to filter what i need to find the problem
because im not exactly sure what the problem is

Steps to reproduce the behavior:

not sure how to reproduce our system has been working for quite a while no problems
we have about 35 agents and about 120k of tickets.

hopefully someone can direct me to the problem to resolve it

thanks

[KINGRULZ]

Have just set a cron to restart zammad-worker every 30 minutes seems to keep the delayed jobs away

[KINGRULZ]

now that the cron is keeping the delayed jobs under control, this error is what i think is causing the worker or scheduler to crash
“scheduler may not run (last execution of Stats.generate 23 minutes ago)”

always showing up at that 23 - 25 minute mark @MrGeneration have you any idea why this stats.generate would cause some problems ?

[MrGeneration]

Restarting the scheduler helps? Well I’d look there if I’d be you.

[KINGRULZ]

Error shows in zammad monitor

what i see in Logs for errors is nothing

[MrGeneration]

If your log files don’t return anything then the process is either killed or busy with other stuff.
Both scenarios can have various reasons. At least the killing part would be logged by your system.

This is as far as I can help.

[dvnkln]

Try updating to 5.3.0.
We had a similiar problem with tons of background jobs while running 5.2.3.

It could™ be that you’re affected by this:

Updating group settings causes huge numbers of delayed jobs

There’s some workaround without needing to update but well, you know. Using the most up-to-date version is always the better choice.

cheers

[KINGRULZ]

Thanks i get the install pretty up to date, and it was fixed with 5.3. The work around i had was restart the workers via cron job every 15 minutes to keep the delay jobs down to allow system to be fairly stable for end users.