Out of the box, when an exception is thrown on production (non debug mode), Flask will log the traceback and and execute the corresponding handler (either custom or default).
In some circumstances, just a traceback is not enough to figure out obscure bugs. Info like the session, the request arguments, cookies, etc can be very handy, and it would be nice to have them logged too.
There are two straightforward ways to do this. You can subclass Flask class and override handle_exception() method:
class MyFlask(Flask): def handle_exception(self, e): # add all necessary log info here log.info("dumping session: %s", session) log.info("dumping request: %s", request) log.info("dumping request args: %s", request.args) return super(MyFlask, self).handle_exception(e)
or, you can subscribe to the got_request_exception() signal, which is emitted inside Flask.handle_exception() method:
def dump_environment(e, **extra): # add all necessary log info here log.info("dumping session: %s", session) log.info("dumping request: %s", request) log.info("dumping request args: %s", request.args) from flask import got_request_exception got_request_exception.connect(dump_environment)
Keep in mind that the information you want to log depends on your needs, session, request and request.args are just examples, you should change it for whatever you think is useful to log in your own environment
This snippet by Paulo Bu can be used freely for anything you like. Consider it public domain.