[GH-ISSUE #911] [QUESTION] How to handle shutdowns/restarts with systemctl #456

New issue

Closed

opened 2026-03-02 05:21:18 +03:00 by kerem · 1 comment

kerem commented

2026-03-02 05:21:18 +03:00

Owner

Originally created by @arjendevos on GitHub (Aug 28, 2024).
Original GitHub issue: https://github.com/hibiken/asynq/issues/911

In production we compile to a single go binary and run it through systemctl. However restarts/stop don't work as they hang forever. I know we have to use kill -TERM etc but we don't know the in our CI?

Is there any docs or examples for an use-case like this?

EDIT:
I now have this:

# Graceful stop
ExecStop=/bin/kill -TSTP $MAINPID
ExecStop=/bin/sleep 10
ExecStop=/bin/kill -TERM $MAINPID

I get these logs:

INFO: Stopping processor
INFO: Processor stopped
INFO: Starting graceful shutdown
INFO: Waiting for all workers to finish...
INFO: All workers have finished
INFO: Exiting

However it seems to hang on Exiting and it never completes?

Originally created by @arjendevos on GitHub (Aug 28, 2024). Original GitHub issue: https://github.com/hibiken/asynq/issues/911 In production we compile to a single go binary and run it through systemctl. However restarts/stop don't work as they hang forever. I know we have to use kill -TERM <pid> etc but we don't know the <pid> in our CI? Is there any docs or examples for an use-case like this? EDIT: I now have this: ```bash # Graceful stop ExecStop=/bin/kill -TSTP $MAINPID ExecStop=/bin/sleep 10 ExecStop=/bin/kill -TERM $MAINPID ``` I get these logs: ```bash INFO: Stopping processor INFO: Processor stopped INFO: Starting graceful shutdown INFO: Waiting for all workers to finish... INFO: All workers have finished INFO: Exiting ``` However it seems to hang on Exiting and it never completes?

kerem closed this issue

2026-03-02 05:21:18 +03:00

kerem commented

2026-03-02 05:21:19 +03:00

Author

Owner

@arjendevos commented on GitHub (Aug 28, 2024):

Oke found the issue, I was running it in a goroutine. Makes sense now I think about it. Changing it to the main thread fixed everything.

@arjendevos commented on GitHub (Aug 28, 2024): Oke found the issue, I was running it in a goroutine. Makes sense now I think about it. Changing it to the main thread fixed everything.

kerem referenced this issue

2026-03-07 22:07:39 +03:00

[GH-ISSUE #456] [FEATURE REQUEST] Is there any way to do some specific action before periodic task enqueue? #1223

kerem referenced this issue

2026-03-07 22:07:44 +03:00

[GH-ISSUE #459] [FEATURE REQUEST] Periodic Tasks support customize cron logger #1226

kerem referenced this issue

2026-03-15 19:47:37 +03:00

[GH-ISSUE #456] [FEATURE REQUEST] Is there any way to do some specific action before periodic task enqueue? #2235

kerem referenced this issue

2026-03-15 19:48:56 +03:00

[GH-ISSUE #459] [FEATURE REQUEST] Periodic Tasks support customize cron logger #2238