On Wed, Sep 16, 2015 at 2:51 PM, Roberto De Ioris <[email protected]> wrote:

>
> > On Wed, Sep 16, 2015 at 2:37 PM, Руслан Закиров <[email protected]> wrote:
> >
> >> Going to try 2.0.11 and report back
> >
> >
> > still fails.
> >
> >        EXTENDED INFO:
> >         uri: /test.html with args:$VAR1 = {};
> >
> >
> > Wed Sep 16 14:41:59 2015 - *** HARAKIRI ON WORKER 1 (pid: 15702, try: 1)
> > ***
> > Wed Sep 16 14:41:59 2015 - HARAKIRI: -- wchan> hrtimer_nanosleep
> > Wed Sep 16 14:41:59 2015 - HARAKIRI !!! worker 1 status !!!
> > Wed Sep 16 14:41:59 2015 - HARAKIRI [core 0] 127.0.0.1 - GET /test.html
> > since 1442403715
> > Wed Sep 16 14:41:59 2015 - HARAKIRI !!! end of worker 1 status !!!
> > Wed Sep 16 14:42:00 2015 - DAMN ! worker 1 (pid: 15702) died, killed by
> > signal 9 :( trying respawn ...
> > Wed Sep 16 14:42:00 2015 - Respawned uWSGI worker 1 (new pid: 15719)
> >
> > Sports INFO [Wed Sep 16 14:42:00 2015] (RU) pid: 15719 Begin request
> > processing /test.html,%20/errors/502.html
> >         EXTENDED INFO:
> >         uri: /test.html,%20/errors/502.html with args:$VAR1 = {};
> >
> >
>
> It looks like nginx is adding a second REQUEST_URI object (uWSGI is not
> able to do it without internal routing rules)
>
> How do you manage the fallback in nginx ?


In the above case it's error_document directive. In the case I started from
two requests was not related to each other - just two request coming from
different users happens to hit the same back-end.

Can you report its configuration ?
>

As I was able to reproduce the problem in a test env. I want to try to
reproduce it without nginx by hitting uwsgi port directly.

-- 
Руслан Закиров
Руководитель отдела разработки веб-сервисов
+7(916) 597-92-69, ruz @  <http://www.sports.ru/>
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi

Reply via email to