Hi,
I'm writing a Python app that uses websockets.
The default way to write an uwsgi app looks like this
def app(env, start_response):
# do something
return ['']
`app` is the callable and gets called for each request, in the case of
websockets, `app` runs as long as the websocket connection is open.
An almost equivalent solution is to create a callable object:
class App(object):
def __init__(self):
pass
def __call__(self, env, start_response):
# do something
return ['']
app = App()
The problem here is in the last line. `app` is instantiated only once
and re-used for every request until the worker is restarted. How can I
make sure `app` is re-instantiated for each new request. I know about
the `--max-requests` options, and setting it to 1 seems to do the trick
(i.e. one App instance per websocket), however, I have the feeling (or
read somewhere) that this option is not really suited for production
servers. Is that true? The reason why I need a fresh `App` instance for
each new request is that my `App` objects accumulate some state during
the websocket session that needs to be reset for each new websocket session.
Any idea how I can re-instantiate a callable App-object on each request?
Cheers,
Bastian
--
Bastian Venthur http://venthur.de
Debian Developer venthur at debian org
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
