Kain, Becki (B.) wrote:

That's what we've been suggesting. They keep saying its an os issue

I don't see any evidence to support that conclusion. From your description, the httpd processes are mostly CPU-bound, and there's not much else running on the system, so the same processes, after getting interrupted for using up the time slice, get put right back onto the CPU again, as they're the only processes which are in the "runnable" process queue. What's their prior experience been, precisely? How many pages/minute is this thing sending out compared to a system they say is running properly? What is the complexity of the scripts, css, etc. on this machine compared to the "properly running" machine which they are making the comparison with? To me, it looks like it's serving VERY complicated web pages, which demand a lot of server-side computing before the HTML goes out over the network port.

But thanks

-----Original Message----- From: Aaron Kulkis [mailto:akulkis00@hotpop.com] Sent: Thursday, January 17, 2008 12:45 AM To: Kain, Becki (B.) Subject: Re: [opensuse] Top/lsof

...

They are web processes though, and that's the point of this box - as a web server,so i don't want to nice their processes, since it's the

Kain, Becki (B.) wrote: point

...

of this box. Thanks

If the purpose of the box is to serve web pages, then I see no problem with httpd processes consuming up to 99% per CPU

Actually, that's an ideal which I doubt you'll ever be able to reach.

If the web-serving is more sluggish than desired, then I would recommend to the web-programmers to take a look at their code, and optimize it for better performance.

-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org