On Tue, 2009-05-12 at 20:34 -0400, Aidan Van Dyk wrote:
> * Joshua D. Drake <jd@commandprompt.com> [090512 19:27]:
>
> > Apache solved this problem back when it was still called NSCA HTTPD. Why
> > aren't we preforking again?
>
> Of course, preforking and connection pooling are totally different
> beast...
>
Yes and no. They both solve similar problems and preforking solves more
problems when you look at the picture in entirety (namely authentication
integration etc..)
> But, what really does preforking give us? A 2 or 3% improvement?
It depends on the problem we are solving. We can test it but I would bet
it is more than that especially in a high velocity environment.
> The
> forking isn't the expensive part,
It is expensive but not as expensive as the below.
> the per-database setup that happens is
> the expensive setup... All pre-forking would save us is a tiny part of
> the initial setup, and in turn make our robust postmaster controller no
> longer have control.
I don't buy this. Properly coded we aren't going to lose any "control".
Joshua D. Drake
--
PostgreSQL - XMPP: jdrake@jabber.postgresql.org
Consulting, Development, Support, Training
503-667-4564 - http://www.commandprompt.com/
The PostgreSQL Company, serving since 1997