Out of about one or two hundred systems that I dealt with during the
last several years I saw a hard drive fail only four times. Once, it
was 840 (or similar size) MB WD Caviar in 1995. Then it was an old
120MB Seagate. The third time it was a 2.5MB 5,1/2" Quantum (don't
remember the exact brand) installed in a Compaq NT "server". DOS
'scandisk' marked 40% of the hard drive as bad sectors. Then it was
SCSI Seagate Barracuda 4.2GB which failed in a PPro CAD workstation
(again, bloody Compaq!) because of a very poor cooling - some
components on its circuitry board fried. Has anyone ever looked inside
the machines that Compaq sells as "servers"? I didn't believe it when
I did. It took three different tools and fingers all scratched up, to
figure out how to take the case off. It consisted of two panels
attached by soft metal screws requiring a hexahonal screwdriver. The
first thing that surprised me was that heat monster PPro 200 CPU had
no fan, only a heatsink, two network cards, that big Barracuda,
everyting "cooled down" by a miserable fan 200W power supply's but the
case was so tightly designed that no air was sucked in it in the first
place. Also, no 3,1/2" bays, only 5,1/2". You have to find that
additional mouting gismo that you install a 3,1/2 drive in and then
put the whole contraption into a 5,1/2" bay. Argh, I don't believe a
company can make it a global policy to deploy exclusively Compaq-made
servers in all offices and branches. :(
Oh BTW there are some lovely and sadistic thing you can do to an old
dying but still half working hard drive.. take it apart, plug it in a
system, power it up and do stuff like formatting, partitioning,
installing an OS and watch error messages caused when you shove
different pins under the heads, jam actuator mechanisms and scratch
the platter surface.
--
Vlad Petersen | <vladimip at uniserve dot com>
#include <disclaimer.h> | *Good pings come in small packets*
Vancouver, BC | Windows: for IQs smaller than 95
SIGSIG -- signature too long (core dumped)