Re: [cobalt-users] More wierd disk errors on Raq 3 - fsck apparently won't fix

[flash22@xxxxxxx]

On Sun, 31 Dec 2000, Chip wrote:

> Hi, folks.
> 
> apparently occurred during an involuntary power interruption to the Raq.
> 
> hda: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
> hda: read_intr: error=0x10 { SectorIdNotFound }, LBAsect=16594750, 

etc...

> If anyone has any ideas before I call Cobalt for a replacement drive (it's 
> still under warranty) please let me know.

I've seen drives crash the heads into the platter when the power goes off
unexpectedly, they aren't supposed to do that, but it happens...and
damages the media ...so yes, it's probably sick, but you need to be
checking some things instead of just naking fsck try to fix something it
obviously can's..

if you get it to boot, run badblocks and see if there are in fact
unreadable blocks on the drive, if so you have to replace it...

also run fdisk -l and make sure the pastition map is in fact correct, 
if it got corrupted, it could be telling the kernel the partition is
smaller than it in fact really is and the system won't ley fsck read past
where it thinks the end of the partition is , and you will get bizzare
errors...

assuming you have the data backed up somewhere, the last resort before
sending it back would be to try erasing it and remaking the filesystem,
but this means building the boot block too, you need the recovery cd to do
this, and it will be a waste of time if the drive in fact has bad spots on
it..


g