Re: i/o errors on SCSI-drive - please help

[Alex Friis <alexfriis mobilixnet dk>]

Once upon a time I wrote:

>A complety Linux-newbe I am.
>I'm running 
>Red Hat Linux release 6.1 (Cartman)
>Kernel 2.2.12-20 on an i686
>I did the starndard server install path.
>
>On the console (?) I now and then get:
>
>> scsi MEDIUM error on channel 0 id 0 lun0 CDB...........
>> i/o error dev 08:05 sector 557328
>> Additional sense indicates unrecovered read error recommend
>> reassignment.
>
>I've been searching for a log somewhere to hold messages like above.
>But I can't find it - if it's really there!
>
>'Reassignment' - what does that mean - do I have to reformat my HD or
>what ?

Thanks for all the interest people gave this posting :-(

I've found out myself: 
*********************
The drive is in fact erranous.
The SCSI-entry-utility, used during boot, enables me to low-level-format
and verify
each of the SCSI-drives. During 'verify' a great deal of bad blocks
where reported and I was suggested to reassign those, but the bad blocks
just kept om coming, so I called for a replacement drive.
Had a amusing - long - time strapping and swapping scsi-drives - now
after a good deal of blood and sweat (no tears !) I know how to 'jump' a
drive to determine its SCSI ID.
BTW when it has to be a bootable drive it must be ID-0 or ID-1 - I
finally found out.

At a certain time I only had a single drive (id-1) installed. The Linux
installer
installs all the wanted RPM's and when nearly finished he dies - I guess
trying
to write the bootstrap to SCSI id0 !!


-- 
med venlig hilsen/best regards
Alex Friis
mailto:alexfriis mobilixnet dk