Nonsense tracebacks (traceback hijacking?)

[aehlke]

I'm using gunicorn-0.16.1.

I have a Django middleware that under special circumstances (whenever the response is a 403), inside process_response, calls traceback.format_exc(). I would expect it to contain Django internals and the process_response method from which it was called. However, it is always the following:

File "/usr/local/lib/python2.7/dist-packages/gunicorn/workers/sync.py", line 34, in run
client, addr = self.socket.accept()

File "/usr/lib/python2.7/socket.py", line 202, in accept
sock, addr = self._sock.accept()

error: [Errno 11] Resource temporarily unavailable

I can't make any sense of this. It's obfuscating a bug I'm trying to investigate since I can't get a useful stack trace. Any hints would be appreciated, thanks!

[benoitc]

did you try with the log level to debug?

[aehlke]

I'll give that a shot, thanks. Are you suggesting I will find useful log items, or that it will somehow fix the traceback.format_exc output? (In the former's case, I'll have to do some extra work to float these items up to my attention, as this only happens a handful of times a day.)

[benoitc]

You should have the full trace.. Let me know about the result.

[benoitc]

duplicate from #514

[aehlke]

I set the logger to debug level last friday and the tracebacks over the weekend still have the same gunicorn traceback. Any ideas? I'm not sure how this is related to #514 actually, besides having the same error.

[aehlke]

Ah I understand #514 better now, thanks.

[aehlke]

Unfortunately I'm still having the same issue after upgrading to 0.17.3. The logger level is back to INFO, DEBUG in Django is False, etc. Not sure what to try now. Thanks.

[sohamnavadiya]

I am also facing same issue in gunicorn.

[tilgovi]

@sohamnavadiya please open a new issue and provide as much detail as you can.