httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Cliff Woolley <>
Subject Re: Threaded and ps->quiescing
Date Thu, 02 Aug 2001 23:13:41 GMT
On Tue, 31 Jul 2001, Greg Ames wrote:

> yeah, I'm aware of the problems that happen when you don't have incoming
> requests.  Do you ever see any problems where we quit serving pages?

I just saw a problem when I tried it a minute ago.  I think Paul might
have already reported this when he reported his trials, saying it worked
until it died, but it took a minute to "ramp up".

Here's what I did: start up httpd with the same threaded parms as I gave
you before, which are nearly the defaults.  Two of the children will
immediately get killed off because of the funky default parms.  But during
that time you can simulate something that happens later during a graceful
restart.  Immediately after starting the httpd, call up an ab -n 50000 -c
50 ... you have to wait for the two children to finish dying off one
worker at a time before the benchmark will spin up to full speed.  Until
that time, you can only handle one connection at a time for each child.


   Cliff Woolley
   Charlottesville, VA

View raw message