I have been posing recently, about the slowness on a new HP ML110 G3 SATA
RAID 1 server I recently had installed. I wrote a program to record read,
write and read/write speed comparisons and I found the new HP server to be
only about 50% of the server I was replacing.
Patrick Farrell pointed me to TSAtest ( thanks Patrick ) which I ran in on
both servers. I was surprised to find that the new server actually preformed
better than the old one, which confuses me. Here's what I found:

Server-> Azus HP ML110 G3
OS NW6.0 SP5 NW6.5 SP5
Processor-> P5 2.4Ghz 2.4 Cel4
Memory-> 2GB 2BG
Hard DD-> 40gb IDE 80 SATA
RAID 1=> Software Hardware

TSATEST mb/min 1200 1730 <==== HP is 46% faster

This indicates (to me) that the new server (HP) *should* be quicker, but it
is only approx 50% of the speed if the old homebuilt. Makes me think that
there is some settings in the new NW6.5 that are incorrect or different.
From this test, I *think* I can rule out the RAID as a possible bottleneck.

If anyone has with thoughts and ideas on how to track the performance
problem, please share them.