httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Marc Slemko <>
Subject Re: process model again...
Date Wed, 08 Apr 1998 13:57:31 GMT
On Wed, 8 Apr 1998, Jim Jagielski wrote:

> Dirk-Willem van Gulik wrote:
> > 
> > 
> > > So I've been continuing with trying to figure out if the current
> > > perform_idle_server_maintenance() code is right.  I'm now convinced that
> > > we no longer need MinSpareServer, but can't give a good mathematical
> > > reason for why.  It just doesn't seem necessary to keep many spare when
> > > the server will exponentially spawn if a load presents itself.
> > 
> > Hmm, though the time needed to spawn a new child is short (and might be
> > short compared to the actual handling) I would expect that this delay
> > is in itself the reason for having MinSpareServer's ready.
> > 
> Hmmm... worse case, let's assume MinSpareServer is 5 and we all
> of a sudden get hit by a bunch of requests such that we are 5 process
> short. With MinSpareServer, Apache will be able to handle these
> requests with "no" delay and then spawn extra processes "in the
> background" as well. Without MinSpareServer, Apache could take
> about 3 seconds to have enough processes to handle the
> extra requests, so there's that delay... Most likely I buggered
> something up in my assumptions, but that's how it looks to me :/

Yes, but the reason for that delay is not necessarily intrinsic to the
process, just a side-effect of the current implementation.  Apache could
spawn servers on demand a lot faster than it does.

View raw message