Grouped jobs firing multiple times

[hardcodet]

Noticed the following issue here in my logs:

1. I'm scheduling a grouped job (concurrency: 1) for execution
2. BullMQ triggers the execution (with quite a delay, even though it was scheduled for immediate execution, and there weren't too many jobs. This may be a red herring though).
3. about 30 seconds later, the job is triggered again

• I looked for similar log entires, and it appears this happens from time to time, and then in clusters.
• May be a load issue.
• No services were restarted during those times
• we're currently running version 5.1.10

Any idea why this is happening?

[manast]

As you mentioned it would be quite useful to understand why the job is delayed. Could there be an issue with the workers? are they really online? A job processed twice is often produced by a job that has been stalled as if the worker has lost connectivity or hung for too long, which makes the job get stalled, which will result in a reprocessing of the same job. Check the job details to see if it has been attempted several times, that would be an indication of this behaviour.