> > The cons side of processes model is not the startup time. It is about
> > kernel resource and context-switch cost. Processes consume much more
> > kernel resource than threads, and have a much higher cost for context
> > switch. The scalability of threads model is much better than that of
> > processes model.
>
> My question here is how much do we really context switch. We
> do quite a
> bit of work for each query, and I don't see us giving up the CPU very
> often, as would be the case for a GUI where each thread does a little
> work and goes to sleep.
Every IO gives up the CPU ! In a threaded model the process could use up it's
CPU timeslice for other clients.
The optimum would be a multi process multi threaded server. But this only starts to
show an effect if there are a lot of clients (thousands).
Andreas