[SWLUG] Re: Intermittent Hard disk problem

[Carter N.]

>I switched the system on and it made a funny noise.
>Then it failed to boot from the C drive.
...
>They were fine. Then I put the drives back into the PC
>and rebooted. The system worked!

I've seen similar problems with two old (6.4GB samsung IDE UDMA 66) drives
as master and slave on the primary channel. One of the drives made a nasty
sound and stopped working for a few reboots (not detected by POST), but
eventually started working fine again.

Removing one drive and running just the faulty one on its own seemed to
help, but the notion that this configuration actually did anything in itself
is probably just superstition.

It failed again on a much later occasion (I think it was the same drive,
dunno), and started working again after a day's 'rest'. Actually, in this
case, I had booted with Tom's root boot disk, and did the mount, chroot, and
lilo waltz to reinstall the boot record for the primary master (this
contains Win98, the slave contains Minix and Linux) -- as mentioned in the
recent post by Tim Shapcott. Once I got them rebooting again, the data and
OS were all intact on all partitions and both drives.

To this day, I don't know what went wrong, and it may well happen again. For
my casual home purposes, it's not worth fitting reliable hard disks so I
keep both well backed-up; Although I never actually lost any data as a
consequence of these failures.

Having said that, at work, given the ratio between the cost of a hard disk
and the likely value of the data on it, any hard disks that show physical
problems are scrapped immediately, regardless of backup status.

By the way, how do you obtain SMART messages?

Hope this helps.

Neil
--------------------------------------------
Neil Carter            Psychology Department
IT Technician    University of Wales Swansea
                       Wales, United Kingdom

http://psy.swansea.ac.uk/staff/Carter/