Replies: 1 comment 9 replies
-
@larsent this is gonna be tricky to figure out, but I have a hunch it's some interaction between our First, can you tell me how you initialize Sentry in your Celery workers and the Second, I'd ask you to disable the cc @antonpirker |
Beta Was this translation helpful? Give feedback.
9 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I'm creating a discussion instead of an issue because I'm not positive this is related to sentry-python, but I'm not sure how to debug further. I'd welcome advice.
I'm running Django + Celery. After upgrading sentry-python from 1.32.0 to 2.3.1, I started getting several (~3-5) errors per day like "WorkerLostError('Worker exited prematurely: signal 11 (SIGSEGV) Job: 3889.')" in my celery tasks.
I'm using Redis as both my task broker and results backend, and running the following
My celery worker's memory doesn't seem to spike, so I think I've ruled that out as a cause.
I used faulthandler to dump the stack, and pasted the output below. You'll notice code from segment's and scout's sdks in addition to sentry here, but these errors also occur without those sdks, so I don't think they are suspects.
After downgrading sentry-sdk to 2.0.1 yesterday, I have not seen this error again. Again, the error started after the 1.32.0 to 2.3.1 upgrade. That's why my best theory so far is that this is somehow related to sentry-python. Do you have any advice for debugging this further? Thanks.
Beta Was this translation helpful? Give feedback.
All reactions