Hello!
I maintain a Django app hosted on Heroku. We use DRF intensively to serve our FE app.
Django version: 2.2.21.
We have an endpoint that works perfectly well 99% of the time, but sometimes, it timeouts.
We use Sentry to monitor and you can find the traces
here.
FYI but maybe it is not useful, this endpoint:
- Has very little processing then starts a background task.
- In the breadcrumb of Sentry, we can see that the REDIS LPUSH command is triggered at the very end of the 20s heroku timeout.
It seems to me that the function calls is going back and forth between
- django/core/handlers/exception.py in inner at line 34
django/utils/deprecation.py in __call__ at line 94
This has been happening for months now and we still do not have a solution.
Any idea as to why this might happen?
And why only in certain cases?
Thank you very much for the help!
Robin Bonnin