From owner-freebsd-questions  Tue Nov 21 11:29:32 1995
Return-Path: owner-questions
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id LAA14831
          for questions-outgoing; Tue, 21 Nov 1995 11:29:32 -0800
Received: from sh1.ro.com (mprevost@sh1.ro.com [205.216.92.5])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id LAA14826
          for <questions@freebsd.org>; Tue, 21 Nov 1995 11:29:28 -0800
Received: (from mprevost@localhost) by sh1.ro.com (8.6.9/8.6.9) id OAA24880; Tue, 21 Nov 1995 14:37:30 -0600
Date: Tue, 21 Nov 1995 14:37:30 -0600
From: "Mike R. Prevost" <mprevost@ro.com>
Reply-To: "Mike R. Prevost" <mprevost@ro.com>
Subject: [SOLVED] 2.1R Very Stable Here
To: FreeBSD Questions <questions@freebsd.org>
In-Reply-To: <Pine.SUN.3.91.951121084022.23222A-100000@garnet>
Message-ID: <Pine.3.89.9511211432.A24329-0100000@sh1.ro.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
Sender: owner-questions@freebsd.org
Precedence: bulk


I had some trouble with FreeBSD on a new news server here.  With the
help of several folks in this list and on usenet, I proved it to be a bad
motherboard.  Got new ASUS 486SP3G motherboard and things are grand.
System has a 3c509, 1 1GB Conner, and 2 2 GB Seagate Baracuddas in it at 
present.  2 More Seagtes on the way.  32 MB RAM.

Just thought I'd post this script that I used to stress test the system.
I left it running all night and it was still running when I got back the 
next morning.  Some filesystems had filled up though.  Not sure why.  But
I ran a similar script on a Linux machine and it die during the night.

The FreeBSD system kept a steady load avg of about 13.5 all nigh with the 
disks grinding like mad.  I left 3 sysstat's running (with pigs and 
iostat and vmstat too).

Anyway, thanks for your help.  I'v been running Linux for a long time 
now.   I'm very impressed with FreeBSD so far.

Here's the quick little I/O stress testing script (/sd1 is the first
Seagate, /sd2 is the second.  / and /usr are on the Conner): 

--- snip snip ---
#!/bin/sh

while true
do
  sync
  sync
  sync
done &

while true
do
  find /usr -exec cat {} > /dev/null  \;
done &

while true
do
  find /usr -exec cat {} > /sd1/tmp \;
done &

while true
do
  find /usr -exec cat {} > /sd2/tmp \;
done &

while true
do
  nice -20 cat /dev/mem /dev/kmem > /sd1/mkmem
  rm -f /sd1/mkmem
  sync
done &

while true
do
  nice -20 cat /dev/mem /dev/kmem > /sd2/mkmem
  rm -f /sd2/mkmem
  sync
done &

while true
do
  nice -20 dd if=/dev/rsd1 > /dev/null
done &

while true
do
  nice -20 dd if=/dev/rsd2 > /dev/null
done &

while true
do
  nice -20 ps auxw > /dev/null
done &

while true
do
  nice -20 vmstat > /dev/null
done &

while true
do
  nice -20 iostat > /dev/null
done &

while true
do
  nice -20 cat /dev/mem /dev/kmem | nice -20 /sbin/md5 | nice -20 gzip -9 | nice -20 gunzip > /dev/null
done &

while true
do
  echo > /dev/null
done &
--- snip snip ---


--- Mike R. Prevost