exited with code 73

[celery]

exited with code 73

[mheppner]

You'll have to isolate the issue and figure out which part is actually causing the exit code. Exit code 73 usually means: can't create (user) output file. However, it's not clear what exactly you're reporting here...are you asking what is causing the strange exit code, or are you asking why celerybeat doesn't delete it's pid?

if celerybeat.pid existence， it return exit code 73,

[mheppner]

Can you clarify why you think that's an error with django-celery-beat?

Regardless, this is a pretty big stack that you haven't given us much info about. What base images are you using? Docker engine version? Celery & Python versions?

How are you stopping the Docker container? If you use kill, that's a hard kill command to the container and Celery beat probably doesn't know about it. Try using stop, which should actually send a command to Celery beat and let it clean itself up.

@mheppner
my dear friend
thank you for reply

i used https://github.com/shipyard/shipyard stop|restart my Docker container,
when is used https://github.com/shipyard/shipyard restart celery_beat container, it return exit code 73, celerybeat.pid existence

i think maybe we can :
celerybeat.pid content is a Unix Time Stamp,
when celery beat running, every 60 seconds, we rewrite celerybeat.pid , write now Unix Time Stamp,

---

so , every time, we running celery beat ,if celerybeat.pid existence,we read celerybeat.pid content, if content more than 60 seconds, we know, last time, celery beat abnormal exit，then we update celerybeat.pid content,

Please forgive me for my poor English

[mheppner]

You'll need to isolate where this is happening. Try launching your docker manually and send a stop command vs a kill command. You still haven't narrowed down what the issue is you're having with this project, or even Celery itself for that matter. This shouldn't be reported as an issue here, you should be going to the mailing list instead to ask for help.

With that being said, your celerybeat.pid file is the process ID, not a timestamp. I've never seen a pid file contain a timestamp before.