Jeff Trawick wrote:
> 
> Greg Ames <[EMAIL PROTECTED]> writes:
> 
> > ...as of Fri Jan 11 15:05:33.  The purpose of this build was to see if the load
> > spikes happen with any build done on FreeBSD 4.5 because of our problems with
> > more recent httpd code.  So far I haven't seen the spikes.  Also, my patch to
> > log cpu per request is running.
> 
> Does the lack of spikes coincide with a lack of unreasonable cpu ticks
> per request?

no, the unreasonable cpu ticks are still there in 2_0_28, so we can't use them
to help debug the load spikes.  Brian P suggested the former might be due to
spin locks.

> Eyeballing a "tail -f weblog" I just see ticks in the 0-3 range, with
> almost no 3's and lots of 0's.  

yeah, something like 98% of requests have less than 3 cpu ticks.

> (But I didn't whip out /usr/bin/perl to do a real check.)

well, I did and I saw a few negative numbers...wtf?!?  It doesn't matter, this
patch isn't helping us now anyway.

Greg

Reply via email to