How to see exceptions in a Flask + gunicorn app?

Question

For the life of me, I can't figure out where my errors are coming from in my Flask app when I run it through gunicorn, as I can't figure out how to get a stack trace displayed.

For example, let's say I have a very simple "Hello, World!" app written in Flask.

import logging
import os
from flask import Flask

app = Flask(__name__)
app.debug = True

@app.route('/')
def hello():
    raise Exception('Exception raised!')
    return 'Hello World!'

if __name__ == '__main__':
    app.run()

If I run this with python hello.py , then all is well, as I get a very useful stack trace:

(venv)142:helloflask $ python hello.py
 * Running on http://127.0.0.1:5000/
 * Restarting with reloader
127.0.0.1 - - [30/Jun/2014 18:52:42] "GET / HTTP/1.1" 500 -
Traceback (most recent call last):
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1836, in __call__
    return self.wsgi_app(environ, start_response)
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1820, in wsgi_app
    response = self.make_response(self.handle_exception(e))
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1403, in handle_exception
    reraise(exc_type, exc_value, tb)
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1817, in wsgi_app
    response = self.full_dispatch_request()
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1477, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1381, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1475, in full_dispatch_request
    rv = self.dispatch_request()
  File "/Users/grautur/code/helloflask/venv/lib/python2.7/site-packages/flask/app.py", line 1461, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/Users/grautur/code/helloflask/hello.py", line 10, in hello
    raise Exception('Exception raised!')
Exception: Exception raised!

However, if I create a Procfile with

web: gunicorn hello:app

and then start the app with foreman start -p 8000 , then I see nothing at all. Just an "Internal Server Error" webpage.

$ foreman start -p 8000
18:56:10 web.1  | started with pid 36850
(...nothing else is ever displayed...)

How do I get my Flask+gunicorn app to show more useful debugging messages? I tried setting app.debug = True (and various iterations thereof), as it seemed like other StackOverflow posts suggested, but it doesn't seem to change anything.

Answer 1

I'm not familiar with Foreman, so not sure if it needs to be configured in any way, but for Gunicorn to log anything, first you need to set up logging for your Flask app.

The simple way to do it would be like the following:

import logging

# Log only in production mode.
if not app.debug:
    stream_handler = logging.StreamHandler()
    stream_handler.setLevel(logging.INFO)
    app.logger.addHandler(stream_handler)

StreamHandler logs to stderr by default. If you want to log to a file, Gunicorn has --access-logfile and --error-logfile options.

For more detailed explanations and a few good ideas about Flask error handling see the docs .

Answer 2

I found that when Gunicorn was being activated with an Upstart script (Ubuntu 14.04 LTS), errors from Flask were being written to /var/log/upstart/YOUR_UPSTART_SCRIPT.log .

No matter what else I tried, eg the solution by @Audrius Kažukauskas, I simply couldn't see any errors or tracebacks anywhere.