What is the most optimum server configuration for OSQA?

Your apache configuration really doesn't make much of a difference until you start getting many concurrent users (i.e. you have multiple outstanding page requests at the same time). The first thing you should do, before you even worry about playing with your apache configuration, is get caching in place to reduce database access. CACHE_BACKEND in your settings file lets you configure your caching. See the Django docs for specifics on how to set this up. You can turn on file-based caching right now and start seeing huge improvements. However, the ideal solution, where possible, is to install memcached for in-memory caching. Caching is basically guaranteed to make a difference an order of magnitude greater than any difference that you could possibly make playing with your apache settings.

Once you've done this, then you can start thinking about apache configuration.

First principles: Use WSGI. Don't use mod_python. It is probably always slower and more demanding of RAM than mod_wsgi. Part of the reason for this is that WSGI allows your application itself to remain resident in memory between requests.

To choose an MPM, you first need to ask yourself if memory is at a premium in your environment. In a very constrained environment like a VPS with 256-512MB of RAM, you probably want to use the prefork MPM with a minimal number of spare processes kept alive.

Where more memory is available, you have some room to stretch out which will in turn let you get better use out of your CPUs. Use the worker MPM (tends to use more RAM but gets better parallelism). Run WSGI in daemon mode, as this will offload some processing from apache in a way that improves apache's performance characteristics for static content like images. Daemon mode also sort of gets you "free" optimization of threading configuration so you don't have to worry as much about tweaking the apache MPM configuration. Drop these directives in your apache site configuration to do this:

WSGIDaemonProcess OSQA processes=5
WSGIProcessGroup OSQA

OSQA can be replaced with some simple name that is meaningful to you, and a good value for processes is number of CPUs (cores) plus 1 (note however that you will get no benefit out of having multiple processes unless you do in fact have multiple requests processing at once on your site, i.e. get a lot of traffic).

Finally, benchmark. Find out what configuration is best on your particular platform.