>>> but this again doesn't restart the mule and the job if it dies.
>>
>> Are you sure the master is enabled ?
>>
>> mules are monitored like the other processes.
>>
>> Can you report what uWSGI says when a mule dies ?
>
> master is enabled
>
> This is my .ini file for testing:
>
> [uwsgi]
> master = true
> harakiri = 30
> http-socket = 127.0.0.1:4001
> plugin = python27
> wsgi-file = %d/app.wsgi
> mules = 6
>
> This is app.wsgi for testing:
>
> from uwsgidecorators import muleloop
> import uwsgi
> import time, random
>
> def worker(s):
> i = random.randint(0, 6)
> x = 1 * 2 * 3 * 4 * 5 / i
> time.sleep(10)
> print 'Worker mule id %d: "%s" running. Result: %d' %
> (uwsgi.mule_id(), s, x)
>
> muleloop(1)(lambda: worker('test-1'))
> muleloop(2)(lambda: worker('test-2'))
> muleloop(3)(lambda: worker('test-3'))
> muleloop(4)(lambda: worker('test-4'))
> muleloop(5)(lambda: worker('test-5'))
>
> from werkzeug.testapp import test_app as application
>
>
> So I see workers start running, printing their results, and as they
> throw ZeroDivisionError exceptions the mules just sit there dormant.
> So the mule processes keep going - I can see them in 'ps', but they
> don't do anything.
>
> This is uwsgi 1.4.4
>
>
So the behaviour of the master is the right one, the problem are mules not
exiting after an exception ?
--
Roberto De Ioris
http://unbit.it
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
