We have a Dell Poweredge 1400SC with 2 pairs of RAID 1 drives running
from the onboard SCSI controller, all the same SCSI channel - 2 x 18GB
Seagate Cheetah and 2x 36 GB IBM. This setup had been fine for maybe 9
months. Then one of the 18gb drives appeared to fail... mirror was
broken and CHKDSK found lots of corrupt files.

I ran Seagate and Dell diagnostics on the failed drive, came up OK so
re-mounted and formatted, and remirrored it. Lasted maybe an hour
before it failed again.

Yesterday I installed a new 36GB Cheetah, using the same SCSI ID (1)
as before, and re-established the mirror. All seemed fine, but then
when I came in this morning it had failed also.

Meanwhile the original "faulty" 18GB drive is running fine in another
system.

Seems to indicate the drive is not the problem - so what else should I
be looking at? I've checked the obvious loose cabling etc.

Is it worth swapping everything to the other SCSI channel to try to
eliminate that as the source of the failure?

thanks

is the onboard controller a raid controller (ie, PERCx/Si) ? Have you
checked firmware and bios updates, and also made sure that any OMSA and OMAM
patches have been installed?

Dell RAID usually works, but it has more than its share of problems. Because
of those problems, we are now back to buying HPaq, even though they are more
expensive. Thier RAID is bulletproof.

NuTs

"tangel" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> We have a Dell Poweredge 1400SC with 2 pairs of RAID 1 drives running
> from the onboard SCSI controller, all the same SCSI channel - 2 x 18GB
> Seagate Cheetah and 2x 36 GB IBM. This setup had been fine for maybe 9
> months. Then one of the 18gb drives appeared to fail... mirror was
> broken and CHKDSK found lots of corrupt files.
>
> I ran Seagate and Dell diagnostics on the failed drive, came up OK so
> re-mounted and formatted, and remirrored it. Lasted maybe an hour
> before it failed again.
>
> Yesterday I installed a new 36GB Cheetah, using the same SCSI ID (1)
> as before, and re-established the mirror. All seemed fine, but then
> when I came in this morning it had failed also.
>
> Meanwhile the original "faulty" 18GB drive is running fine in another
> system.
>
> Seems to indicate the drive is not the problem - so what else should I
> be looking at? I've checked the obvious loose cabling etc.
>
> Is it worth swapping everything to the other SCSI channel to try to
> eliminate that as the source of the failure?
>
> thanks

"nut cracker" <(E-Mail Removed)> wrote in message news:<uUdwZr$(E-Mail Removed)>...
> is the onboard controller a raid controller (ie, PERCx/Si) ?

No, software RAID

> Have you
> checked firmware and bios updates, and also made sure that any OMSA and OMAM
> patches have been installed?

Yes - there are some available, but I haven't installed them (yet).
I'm a believer in "if it ain't broke...." It's been running in this
configuration for 9 months without problems, so I doubt the problem is
related to these updates. However now that it IS broke, and absent any
other more likely explanations, I will do the updates.

> > Is it worth swapping everything to the other SCSI channel to try to
> > eliminate that as the source of the failure?

Is this a worthwhile idea?

Trevor

Pop for the hardware RAID controller. That will make a lot of differences,
and take the load of the raid operations off the CPU. You can find lots of
controllers on ebay, from dell, that work in that model for very little
money.

Give it a look,

NuTs

"tangel" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> "nut cracker" <(E-Mail Removed)> wrote in message
news:<uUdwZr$(E-Mail Removed)>...
> > is the onboard controller a raid controller (ie, PERCx/Si) ?
>
> No, software RAID
>
> > Have you
> > checked firmware and bios updates, and also made sure that any OMSA and
OMAM
> > patches have been installed?
>
> Yes - there are some available, but I haven't installed them (yet).
> I'm a believer in "if it ain't broke...." It's been running in this
> configuration for 9 months without problems, so I doubt the problem is
> related to these updates. However now that it IS broke, and absent any
> other more likely explanations, I will do the updates.
>
> > > Is it worth swapping everything to the other SCSI channel to try to
> > > eliminate that as the source of the failure?
>
> Is this a worthwhile idea?
>
> Trevor