Hi,

I've removed PHP, turned Apache to worker, changed wsgi daemon to
processes=4 threads=75... and omg. It's HUGE difference. :O Really.
>From 60-70 requests currently being processed Apache goes down to
10-15. Thanks for that blog note! Link should be in some visible place
on mod_wsgi home page! :)

On 23 Gru, 22:14, Szymon <[email protected]> wrote:
> On 23 Gru, 22:00, Graham Dumpleton <[email protected]> wrote:
>
> > 1. Are you using separate media server for static files?
>
> Nope. Static content is served from same server, but using nginx.
>
> > 2. Are you using a front end proxy to Apache/mod_wsgi like nginx?
>
> No.
>
> > 3. Are you using prefork or worker MPM?
>
> Prefork.
>
> > 4. Are you running other web application on same Apache using mod_php
> > or other systems?
>
> Yes, there is one PHP app, but it's not busy at all. But if it's make
> difference I can resign from that app at all.
>
> > 5. Is your Python web application running in embedded mode or daemon mode?
>
> Daemon mode.
>
> > 6. Have you read:
>
> >  http://blog.dscpl.com.au/2009/03/load-spikes-and-excessive-memory-usa...
>
> No. :) But I'll do. :)
>
> > It is hard to give any advice without knowing. For all we know, you
> > could just have configured your current system wrongly and if tweaked
> > it could work fine.
>
> Here is my configuration:
>
> WSGIDaemonProcess app user=app group=app processes=30 maximum-
> requests=10000
> WSGIProcessGroup app
>
> And my prefork config:
>
>  StartServers          5
>  MinSpareServers       10
>  MaxSpareServers      25
>  MaxClients          250
>  MaxRequestsPerChild   0
>
> Keep-Alive is turned off. I'm on FreeBSD if that's make difference.

--

You received this message because you are subscribed to the Google Groups 
"modwsgi" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/modwsgi?hl=en.


Reply via email to