Hi Graham and others
On 25.02.2013 10:23, Graham Dumpleton wrote:
What are you setting MaxClients to in Apache? Provide all the MPM settings you
have set for prefork MPM. Eg: settings for:
In file: /etc/apache2/server-tuning.conf
14: StartServers 10
17: MinSpareServers 10
20: MaxSpareServers 20
23: ServerLimit 150
26: MaxClients 150
29: MaxRequestsPerChild 10000
62: KeepAlive On
69: MaxKeepAliveRequests 100
75: KeepAliveTimeout 15
More configs needed?
How long does it take the WSGI application to respond to a SOAP request?
About 1 to 5 seconds.
It is a machine-machine communication (soap). The client creates about 40k
requests. I don't know how many
come in parallel. But more than our server can handle.
You either have MaxClients set to some ridiculously small value, your SOAP
requests take an excessive amount of time to
respond and you aren't providing enough capacity, or your WSGI application is
locking up completely and never responding.
It is no problem if our server can't respond fast to the soap requests.
My problem is that one wsgi app consumes apache workers and other wsgi apps
don't get any.
Our server has 24 CPUs. Up to now I was happy with "WSGIDaemonProcess .... processes=N". For example, if I set N to 5,
then I know that the soap app won't use more then 5 CPUs. This leaves enough capacity to the other (more important,
since interactive) wsgi apps served by the same apache.
Also, is your WSGI application thread safe?
I guess it is not thread safe. We only use prefork.
Thomas Güttler
--
Thomas Guettler, http://www.thomas-guettler.de/
E-Mail: guettli (*) thomas-guettler + de
--
You received this message because you are subscribed to the Google Groups
"modwsgi" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/modwsgi?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
