[Liverpool] Disk read failure issue: assistance needed.

[Dave Brotherstone]

This doesn't look good.  At best, you have a couple of rough bits on your
drive (possible) or some incorrect parameters in the bios (unlikely), at
worst, your drive is on its way to hard drive heaven.  Firstly, you need to
identify which drive it is - the device as listed below is sdc2.  So if you
type "dmesg" (without quotes) you'll get your kernel messages - scrolling
through this should tell you which drive is sdc2.  Do you know which
filesystem is installed on sdc2?  If not, then type "mount", and it should
somewhere say something like...

/dev/sdc2 on / type ext3

It's the last bit you're interested in.  It could be ext2, ext3, XFS,
reiserfs or several others.  Then you can run a fsck (FileSystem ChecK) on
that drive once you're logged in as root in single user mode (which is
probably what has been offered to you).

I'm no expert on fsck, but I think you do :

fsck -n /dev/sdc2
to check the disk.  If the disk was ext2 or ext3 (I think, never had ext3 so
can't be sure), it should tell you how to fix the errors (ie. what options
to run fsck with) in order to try to fix it.  If it's reiser fs, then you'll
need to run fsck -n --fix-fixable /dev/sdc2.

However, if there are bad-blocks detected and fixed, be aware that it's
usually a sign that you're going to get more.  If it can't fix them (ie.
mark them as bad and never use them again), then I think you're stuffed,
short of a low level format to mark them at the disk level as bad.

Good luck, and hope this helps.

Dave.

On 01/04/06, Iain MacGranthin <iaingmacg at mac.com> wrote:
>
> I have SuSe 10.0 running on an AMD64 3500 based system with 3 SATA
> drives and 1 ATA drive.
>
> I generally use it for DVB recording (because of stability), media
> conversion and media storage.
>
> Just in the past few days I have received the following disturbing
> reiterative output to screen during reboot after a series of error
> messages:
>
> Additional sense: Unrecovered read error - autoreallocate failed
> end request:I/O error, dev sdc, sector xxxxxxx
> Buffered I/O error on  device sdc2, logical block xxxxxxxx
> ata3: status 0x51 {DriveReady SeekComplete Error}
> ata3: error: 0x40 {Unrecoverable Error}
> SCSI Error: <2 0 0 0> return code 0x8000002
> sdc: Current: sense key: Medium eror
>
> This is followed by instructions to manually fix before rebooting,
> and I am allowed to log into root.
>
> Therea re more error messages, but as these were iterative, they were
> the ones I could copy straight away: I can run through the boot up
> again and try for more information if necesssary...
>
> Now this is tough for me as I am getting into Linux from the shallow
> end, building from source code and using Yast2 to update, without
> having too much to with terminal input and system maintenance type
> knowledge, but moving on as I get things working for me.
>
> Well it looks like that nice easy learning curve got a spanner in its
> works.
>
> How do I start addressing this problem:
> Is there something I can use to analyse and repair the disk?
> Do I have to change a file to exclude that disk to get Suse to reboot
> into the nice cosy KDE I am used to. If so, will I be able to recover
> the disk?
> Is there another forum I should be talking to?
>
> Should I decide to reinstall, how would I be able to access the disk
> in question - could I perform repairs/data recovery that way?
>
> I really am in the dark here, so any help/pointers would be a start...
>
>
> _______________________________________________
> Liverpool mailing list
> Liverpool at mailman.lug.org.uk
> https://mailman.lug.org.uk/mailman/listinfo/liverpool
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.lug.org.uk/pipermail/liverpool/attachments/20060402/41319a72/attachment.html