Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 9 Feb 2005 21:55:28 -0400 (AST)
From:      "Marc G. Fournier" <scrappy@hub.org>
To:        freebsd-questions@freebsd.org
Subject:   Dual-Xeon vs Dual-PIII ... Dual-PIII actually better?
Message-ID:  <20050209214335.E94338@ganymede.hub.org>

next in thread | raw e-mail | index | archive | help

I've had a discussion going on talking about performance issues of one of 
my servers, and right now the only thing that I've got to "work with" is 
the difference in CPUs ... I had started it off thinking it was a software 
RAID issue, but looking at my Dual-PIII, its not exhibiting near as much 
load, and the only difference between the  two configs is drive sizes 
(Dual-PIII has 36G Seagate Drives, the Dual-Xeon has 73G ones) and the 
CPUs ... even the operating system is within days of each other, so either 
I hit a bad 'cvsup day' for the Dual-Xeon, or I'm missing something as far 
as Dual-Xeon's is concerned ...

vmstat 5 on both machines shows >50% idle CPUs on the Dual-PIII, while the 
Dual-Xeon shows >90% system busy ... if it were vinum related, I'd expect 
that they would both be about as busy on the system side ...

First question ... is there some way of getting 'finer' data on system 
usage?  What is using up 99% of the %CPU, when it happens?  syscalls/sec 
don't seem to 'jump' much when that happens, hovering around the same on 
both servers (between 2k and 4k / sec ...

I'm going to be doing an OS upgrade on that machine over the next couple 
of days, to see if maybe I did just get a 'bad kernel', but if someone can 
suggest something that I can monitor/look at to determine where the sys 
cpu is being sucked up, that would be appreciated ...

thanks ....



----
Marc G. Fournier           Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org           Yahoo!: yscrappy              ICQ: 7615664



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050209214335.E94338>