Re: [ILUG] Results of software RAID testing last night

From: Paul Jakma (paulj at domain alphyra.ie)
Date: Fri 10 Aug 2001 - 16:55:10 IST


On Thu, 9 Aug 2001, Wynne, Conor wrote:

> External Digital Storage Unit, with 6 Hot-plug SCSI 3 HDD's

bastard..

> My question is :
> Is this because of very slow processor and low memory conditions of the box
> itself?

no.

> Or is it more likely that I have a failed drive in the array?

failed: no. that'd be picked up by the controller.

bad blocks on the drive: possible, but SCSI drives are usually good
about picking this up and at the very least reporting it to the OS.

> dodgy. I will need to low-level format the drive itself and try
> again to be sure.

good idea.

Also, check your memory/CPU. silent corruption is far more likely to
be due to RAM than the md layer.

run memcheck86 for a day or so and i believe there was also talk on
this list of a linux 'burn-in' utility a while ago.

> the FASTSCSI controller is also part of the problem due to its limited
> transfer rate. If anyone has a faster - cheap - scsi controller that they
> are willing to part with, please let me know.

Your bottleneck is far more likely to be the 200MHz CPU at the moment
with RAID.

(though a second controller so that you could have 3 disks per
controller would help).

> Anyone know good RAID tools for linux? Is online spare supported in software
> raid?

possibly. check docs for /etc/raidtab.

> CW

--paulj



This archive was generated by hypermail 2.1.6 : Thu 06 Feb 2003 - 13:11:33 GMT