Subject: Re: daily scripts...
To: None <current-users@NetBSD.ORG>
From: Danny Thomas <D.Thomas@vthrc.uq.edu.au>
List: current-users
Date: 07/07/1996 09:36:56
>On 06 Jul 1996 04:24:35 -0400
> mycroft@MIT.EDU (Charles M. Hannum) wrote:
>
> > I heard a rumour that there was a failing disk on vangogh at one
> > point, and the nightly fsck(8) helped them notice problems earlier.
> > I'm not sure I believe this, or that it's relevant, but `whatever'.
>
>The daily fsck helped a friend of mine find a failing disk, too.

While a lot of drives I've seen come from the fasctory with read and/or
write bad sector remapping disabled, I believe the NetBSD SCSI drivers turn
these on. So would a check for an increase in the number of grown defects
be better than fsck which only finds side-effects, and then only in a
subset of the disk? Ideally you'd want to have something regularly subject
every sector to a read/verify and perhaps write test looking for cases
ECC/retries were invoked.

cheers,
Danny Thomas  <D.Thomas@vthrc.uq.edu.au>