Re: Could this be a zero-length partition?

On 2009-01-21, Günther Schwarz <strap@xxxxxx> wrote:
Vahis wrote:

I ran fsck and e2fsck, with same result:

e2fsck: Attempt to read block from filesystem resulted in short read
while trying to open /dev/sdb1
Could this be a zero-length partition?

The drive is a SATA Barracuda, not very old.

Seagate has currently a severe and nasty problem with their firmware on
some SATA drives. Among these are e.g.

ST3500320AS
ST3640330AS
ST3750330AS
ST31000340AS

Is your drive on this list (ls /dev/disk/by-id/)? If so, good luck with
your backups ;-)


ST3500630AS is the one that "disappeared"

I did all the tricks that I know of in this kind of situation with no
luck at all.

Then there was this new kernel so I thought OK, I'll update the kernel
and since I'll reboot anyway, I'll take that drive out for further
examination on some other machine.

Then I still just decided to reboot once before anything else.
And there it was again!

It's back and the file system (ext3) is reported clean.

During the boot I saw something like (approximately):
"ATA4 is slow to respond, be patient"
several times before grub.

But there were no further errors (none that I could see)
This is BIOS, right? Can those messages be scrutinized somehow
afterwards?

Vahis
--
http://waxborg.servepics.com
"Too bad all the people who know how to run this country
are busy running taxicabs or cutting hair."
George Burns
.

Relevant Pages