+1 faced with the same issue, nothing in the logs

+1 restarting the service seems to solve it but only temporarily

Same issue here. I'm also only getting one task from /api/tasks. Guessing that is related.

+1

I noticed that this issue appears when we deploy our app and celery service restarts

Issue is reproduced in the following case:

1. reach flower max_tasks (by default 10000)
2. restart celery

I'm also able to reproduce like that. I reduced max_tasks to 10 and added a do-nothing periodic task with celery beat to run every 100,000 microseconds to make it easier to test.

+1 this is number one priority for me at the moment

Is this issue still active? I see a weird behavior where tasks stop updating in the UI, and the workers all show offline, but the UI is still "working" and the api still seems to work and tasks are still running on the workers. Not sure if it's the same issue?

This issue should still be active. I am experiencing this bug currently.

as @AdrianLC said, restart kill -SIGINT pid can temporarily solve it.

I'm having the same issue, and a bunch of others. It's a shame that the best celery monitoring tool breaks down after mere 10000 tasks.

We're facing the same issue too. Does anyone have any solutions yet?

After running Flower for a while on Prod, the number of messages exceeds far from the default max_tasks config(200k succeeded messages compares to 10k), Flower show only ONE latest processed task.