How Sentry track their own bugs?


#1

I wonder how Sentry track their own bugs. If you (Sentry) use your own dashboard and tools for tracking, how you can avoid flows that may cause infinite submission?

For example: A bug in your server submission flow -> It’s caught and reported to your servers -> The bug appears again the submission flow -> It’s caught and reported to your servers -> And so on…

Excuse my curiosity :slight_smile:
Thanks.


#2

I wonder, is this post off-topic? I believe the answer can make a client trust your service more by understanding how your process and implementations work.

Waiting for your prompt response :slight_smile:


#3

Hi @john,

I can answer the question about avoiding infinite recursion during event processing: Inside of Raven.capture*, we walk up the current stack and check if the code there is anywhere near the code of api/X/store. If that’s the case we just throw away the event.

We additionally use a few dashboards other than our own, but I currently don’t know what those are.