There must be something here I don't understand then.
If the parent will close the connection and do a long running task, and the
child will return the HTTP response, will the child become a regular worker
of the master process and will keep serving requests as usual?
Is this the way you usually do things:
import uwsgi
import sys
def debug(s):
sys.stderr.write(s + '\n')
def application(environ, start_response):
output = 'OK!'
if 'grunt' in environ['REQUEST_URI']:
if uwsgi.grunt(): *# note I didn't pass True this time!*
debug("worker")
output = 'Grunt!'
else:
debug("parent")
long_running_job()
sys.exit(0)
start_response('200 OK', [])
return [output]
On Thu, Aug 25, 2011 at 6:40 PM, Roberto De Ioris <[email protected]> wrote:
>
> Il giorno 25/ago/2011, alle ore 11:51, Yaniv Aknin ha scritto:
>
> > Hi,
> >
> > On a vanilla Ubuntu 11.04 machine, I built uWSGI 0.9.8.5 and tried
> running the following command line:
> > uwsgi --http 0.0.0.0:8080 --module test --grunt 1 --master 1
> >
> > Where in test.py I had simply:
> > import uwsgi
> > import sys
> > def debug(s):
> > sys.stderr.write(s + '\n')
> >
> > def application(environ, start_response):
> > output = 'OK!'
> >
> > if 'grunt' in environ['REQUEST_URI']:
> > # passing True to grunt() so the parent will keep serving the
> request
> > if not uwsgi.grunt(True):
> > debug("parent")
> > output = 'Grunt!'
> > else:
> > debug("worker")
> > # calling disconnect in the worker to disassociate the worker
> from the request
> > uwsgi.disconnect()
> > debug("disconnected")
> > return
> >
> > start_response('200 OK', [])
> > return [output]
> >
> > This causes a segmentation fault after printing "worker" and before
> printing "disconnected" on Ubuntu 11.04, but works fine on OSX. The segfault
> occurs in natty's eglibc's fclose(), called from
> plugins/python/uwsgi_pymodule.c:py_uwsgi_disconnect().
> >
> > Of course, if I won't call uwsgi.disconnect() in the worker, the segfault
> will not occur - but I thought that this will leave the worker keeping the
> connection with the webserver busy and will not let the parent finish the
> request and move on to other business.
> >
> > Before I dig deeper into the segfault, is this something I'm doing wrong
> with the grunt() and disconnect() APIs?
>
>
> I always use the inverse logic, so it is the parent closing the connection
> and the grunt continuing the job
>
> By the way a segfault is always a bug, so i will investigate on this
>
>
> --
> Roberto De Ioris
> http://unbit.it
> JID: [email protected]
>
> _______________________________________________
> uWSGI mailing list
> [email protected]
> http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
>
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
