Job scheduler getting unexpectedly removed

[gee4vee]

Hi, my code is creating a job scheduler which is getting unexpectedly removed after the first delayed job is created. I can tell because I'm responding to the removed event for the delayed job event but I'm never calling removeJobScheduler. Are there any conditions in which a job scheduler would get removed on its own?

I'm on BullMQ v5.56.4 and Redis v7.2.7.

[manast]

No, the job scheduler will not be removed by its own.

[manast]

The job scheduler or the delayed job associated to the job scheduler? If it is the job scheduler that disappears that is extremely weird... the best would be if you can reproduce it so that we can look into it.

[gee4vee]

@manast Both. I can reproduce consistently only when running in our Kube cluster. Hence I’m trying to get more logs from BullMQ.

[gee4vee]

@manast So is there no way to get more debug logs since I cannot reproduce locally? Trying to see how far I can debug this. If I can't solve it, I'll have to abandon BullMQ for another library which I really don't want to do. It seems to be very stable locally and very easy to use, so I just need to figure out what's happening in Kubernetes that is different.

[manast]

We support open-telemetry, (https://blog.taskforce.sh/how-to-integrate-bullmqs-telemetry-on-a-newsletters-subscription-application-2/) I do not know if that will help you though. I do not think this is a bug in BullMQ, as this is the first time somebody reports an issue like this, if you cannot reproduce it then yes, by all means try other library that works better for you.

[gee4vee]

@manast I appreciate that, I'm just trying to avoid going to another library since the other options are not going to provide all the features we need. Can you point me to some areas of your codebase where I could look at to see where job schedulers could be removed? I'm trying to determine what could be happening in the pods that could be causing the scheduler to be removed, or something the pods are doing to cause that, e.g. initialization or closing or something. Could our pods be doing something to the Redis DB to cause the scheduler to be removed unexpectedly? Anything at all that could help would be appreciated.