"LockNotOwnedError: Cannot extend a lock that's no longer owned" When laptop goes to sleep.

[Cediddi]

Hi,

I'm running redis and redbeat in docker containers. I've set celery_beat_max_loop_interval=5 and redbeat_lock_timeout= celery_beat_max_loop_interval * 5 which means it takes 25 seconds for redis to release the lock and every 5 seconds, the ttl gets extended.

My problem happens when I let my laptop go to sleep for more than 25 seconds (go grab a coffee) and when I reopen my laptop's lid, redbeat freaks out. It says I cannot release a lock that's no longer owned because the process is shutting down because it cannot extend a lock that is no longer owned (or existing)

I think this is not a bug but a design decision. Then what should I do? Should I set --restart=on-failure on docker?

[hotwater9]

I run into the same problems as you, I do the same on my server. But when they work correctly for a while, the problem randomly arises. Nothing happens to the analysis log, as if the lock really expired and disappeared, but shouldn't it be renewed all the time?
So I was wondering if you found the problem?

[Cediddi]

I couldn't find a solution, but I increased celery_beat_max_loop_interval=30 in addition to redbeat_lock_timeout: int = celery_beat_max_loop_interval * 5. I've also made the containers restart on failure with docker compose. On kubernetes, the same.

[hotwater9]

It means that you solved the problem by shortening the polling interval of the beat service, ok，which is the only solution I have found so far