On Wed, 2007-03-28 at 09:30 -0600, adam fisher wrote:
> I just brought up a new web server on fedora core. The load aver appears to 
> grow to the size of the MaxClients which is 150. If I lower it then the load 
> average lowers. 
> 
> When I run ps -aux | grep httpd, I have almost every httpd process in 
> uninterruptible sleep. Are these persistent connections that just aren't 
> being closed or is it something else? 

Reading ahead, it looks like NFS has already been discovered as the
culprit, so I'll keep my message short. Any time you see a lot of
processes in uniterruptible sleep, that probably means you're
bottlenecking on I/O. You need faster hard drives, a fatter network
pipe, or similar.

-- 
Stuart Jansen              e-mail/jabber: [EMAIL PROTECTED]
                           google talk:   [EMAIL PROTECTED]

"However beautiful the strategy, you should occasionally look at 
the results." -- Winston Churchill

Attachment: signature.asc
Description: This is a digitally signed message part

/*
PLUG: http://plug.org, #utah on irc.freenode.net
Unsubscribe: http://plug.org/mailman/options/plug
Don't fear the penguin.
*/

Reply via email to