appsignal/appsignal-python

Research and document how to use Django with AppSignal in production

unflxw opened this issue · 2 comments

unflxw commented

Intercom conversation: https://app.intercom.com/a/inbox/yzor8gyw/inbox/admin/5246522/conversation/16410700274092#part_id=comment-16410700274092-21909541736

This issue should be in either appsignal/test-setups or appsignal/appsignal-docs. So it is here instead. 🙃

Our current instructions tell you to add appsignal.start() to manage.py, which is only used during development. In production, Django recommends deploying it as a WSGI or ASGI server. When doing so, manage.py is never called.

Django automatically creates wsgi.py and asgi.py files for your Django project (see our existing test setup) so adding appsignal.start() there might work? We should find out how it works, add test setups for it, then document it.

To do

  • Create test setups for Django: appsignal/test-setups#190
    • using a WSGI server
    • using an ASGI server
  • Document how to add appsignal.start() when using Django with a WSGI/ASGI server for production use
unflxw commented

Another affected customer: https://app.intercom.com/a/inbox/yzor8gyw/inbox/conversation/16410700293724

Bumping this to the top of to-do since I'm on support this week and it shouldn't be a tough one.