Subject: Re: RaidFrame Raid5 question
To: Greg Oster <oster@cs.usask.ca>
From: Manuel Bouyer <bouyer@antioche.eu.org>
List: port-alpha
Date: 08/29/2005 20:33:20
On Mon, Aug 29, 2005 at 09:04:10AM -0600, Greg Oster wrote:
> sp4rc writes:
> [snip]
> > 
> > so one can assume the harddisk is damaged, but if i disklabel sd6 and
> > set the a partition to fstype 4.2BSD and then create a newfs (newfs -O 1
> > /dev/sd6a) on it, a fsck -fy /dev/sd6a doesn't show up any kind of
> > error. So what's going on here ?! 
> 
> newfs doesn't touch all the blocks on a disk.  And in your case, it's
> not touching the part of the disk that's having problems.
> 
> Try something like this:
> 
>   dd if=/dev/zero of=/dev/rsd6a bs=1m seek=1
> 
> to do the testing.  It should generate the same error that you saw
> with 'raidctl -v -i'.

The error was on a read. Writing to the bad sector may auto-relocate it.

-- 
Manuel Bouyer <bouyer@antioche.eu.org>
     NetBSD: 26 ans d'experience feront toujours la difference
--