I was using gevent and add offload-threads to solve this, I was using
gevent and uwsgi from git repos on 1.4.4 ad latest rc for gevent.
Config was in js (sorry for the format this is automatically generated):
{"uwsgi": {"plugins": ["python27"], "gid": "nuage", "env":
"DJANGO_SETTINGS_MODULE=demo.settings", "pythonpath": ["/home/nuage/code",
"/home/nuage/code/demo"], "workers": 4, "socket": "192.168.2.2:8000",
"autoload": true, "log-date": true, "gevent": 16, "master": true, "eval":
"import django.core.handlers.wsgi\napplication =
django.core.handlers.wsgi.WSGIHandler()\n", "vacuum": true, "home":
"/home/nuage/.virtualenvs/nuage", "no-orphans": true, "log-master": true,
"static-map": ["/static/=/home/nuage/code/demo/demo/static/"], "uid":
"nuage"}}
On Tue, Jan 8, 2013 at 1:54 AM, Roberto De Ioris <[email protected]> wrote:
>
> > I'm using uwsgi 1.4.4.
>
> You should report your config and the startup log lines.
>
> Are you using plain async mode or some coroutine plugin (like gevent) ?
>
>
> >
> > On Mon, Jan 7, 2013 at 3:08 PM, Jorge Cardona
> > <[email protected]>wrote:
> >
> >> Hi,
> >>
> >> I'm suffering this with sendfile using async, is there any common fix
> >> for
> >> this?
> >>
> >> Bye.
> >>
> >>
> >> On Mon, Oct 1, 2012 at 7:47 AM, Robert Klep <[email protected]>
> >> wrote:
> >>
> >>> Roberto De Ioris <[email protected]> wrote on Mon Oct 01 2012 at
> >>> 14:43:16:
> >>> >
> >>> > Can you check if all of the 'broken' requests are served via
> >>> 'sendfile()'
> >>>
> >>> No, they aren't (same INI file as before, with logging enabled):
> >>>
> >>> [pid: 94789|app: 0|req: 1/1] 127.0.0.1 () {26 vars in 536 bytes} [Mon
> >>> Oct
> >>> 1 14:45:27 2012] GET
> >>>
> /biogids/backend/get_records/?rev=0&bitarray=AA%3D%3D&calback=Ext.data.JsonP.callback1&_dc=1349086271113
> >>> => generated 1148108 bytes in 3496 msecs (HTTP/1.1 200) 1 headers in 59
> >>> bytes (1 switches on core 0)
> >>> [pid: 94792|app: 0|req: 1/2] 127.0.0.1 () {26 vars in 536 bytes} [Mon
> >>> Oct
> >>> 1 14:45:34 2012] GET
> >>>
> /biogids/backend/get_records/?rev=0&bitarray=AA%3D%3D&calback=Ext.data.JsonP.callback1&_dc=1349086271113
> >>> => generated 1148108 bytes in 3294 msecs (HTTP/1.1 200) 1 headers in 59
> >>> bytes (1 switches on core 2)
> >>>
> >>> These two requests (made to the Django app) resulted in responses of
> >>> 1022901 and 1144793 bytes, respectively.
> >>>
> >>> -- robert
> >>>
> >>> _______________________________________________
> >>> uWSGI mailing list
> >>> [email protected]
> >>> http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
> >>>
> >>
> >>
> >>
> >> --
> >> Jorge Eduardo Cardona
> >>
> >> [email protected]
> >> [email protected]
> >>
> >> cardona.co
> >> github.com/jorgeecardona
> >>
> >>
> >
> >
> > --
> > Jorge Eduardo Cardona
> >
> > [email protected]
> > [email protected]
> >
> > cardona.co
> > github.com/jorgeecardona
> > _______________________________________________
> > uWSGI mailing list
> > [email protected]
> > http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
> >
>
>
> --
> Roberto De Ioris
> http://unbit.it
> _______________________________________________
> uWSGI mailing list
> [email protected]
> http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
>
--
Jorge Eduardo Cardona
[email protected]
[email protected]
cardona.co
github.com/jorgeecardona
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
