Hey all :)
I'm running into an issue with uwsgi-1.1.1 on Mac OS X 10.7.3, using Python.
Current setup (narrowed down from what I actually want, to make tracking down
this issue easier):
- one vassal in emperor mode: 40 threads, 5 processes, master == true.
The vassal runs the simplest WSGI-app from a module:
def application(environ, start_response):
status = '200 OK'
response_headers = [('Content-type','text/plain')]
start_response(status, response_headers)
return ['Hello world\n']
What I want: use @timer to periodically run a function.
However, whenever I add a @timer (either by decorator, or manually using
uwsgi.add_timer), things start crashing:
Tue Mar 27 18:59:04 2012 - uWSGI worker 3 screams: UAAAAAAH my master died, i
will follow him...
It doesn't seem to matter where I initialize the timer: in the same module as
the application, or in a different module using 'pyimport', same issue occurs.
The fine manual states: "The best way to register signals, is defining them in
the master". Is there a way to explicitely do that? BTW, changing 'processes'
to 1 fixes the issue too, so it seems that the workers are caught by surprise
when they receive the signal.
My actual goal: I'm running multiple Django installs in Emperor mode, which in
itself works great. However, I want to implement Django's "reload when a file
changes" using a @timer, which bombs (an alternative from using a timer is to
use Django's request_started hook or something to check for changes).
-- robert
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
