Chacham's Journal: Hard drive died.... again

Ugh. My HD just gave me the infamous:

EXT3-fs: can't read group descriptor 0
hda: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
hda: read_intr: status=0x41 { Uncorrectable Error }, LBAsect=73, sector=10
end_request: I/O error, dev 03:01 (hda), sector 10
EXT2-fs: unable to read group descriptors
Kernel panic: vfs: Unable to mount root fs on 03:01

If I use a boot disk I get pretty much the same, except it's predicated with:

hda: read_intr: status=0x51 { DriveReady SeekComplete DataRequest Error }
hda: read_intr: status=0x40 { Uncorrectable Error }, LBAsect=73, sector=10
end_request: I/O error, dev 03:01 (hda), sector 10

*sniff*

Maxtor drive. X froze. I had disabled Xscreensaver, so that wasn't the issue. Couldn't even ping it. So, I cycled it and got the error. This seems to have happened last time when I ran Diablo under WineX and accidentaly shook the powercord. Seems to me that the HD died when powering off. Is the motherboard faulty? Or are Maxtors really that bad?

I don't think so

[Twirlip of the Mists]

I've owned many Maxtor hard drives, and I've never had a problem with any of them. I don't mean to sound like a troll-- although the accusation is not unfamiliar to me-- but have you considered the possibility that this is a software problem? Maybe there's nothing wrong with your hard drive. Maybe Linux is to blame.

Just food for thought.

Re:I don't think so

[Chacham]

Unlikely that it is to blame. For the simple reason that this is the 'mount' command. That is common to almost all Un*x flavors.

Anyway, the newsgroups show the error to be a hard drive failure. Even if Linux caused it to fail, that is pretty bad that a hard drive can't survive being turned off whilst running.

Re:I don't think so

[Twirlip of the Mists]

I was actually thinking about the filesystem, not the command that resulted in the error. But like I said, it was just a thought.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:Welp

[Chacham]

Did you try fscking the drive? If it is a filesystem problem, fscking the drive should fix the problem.

No. I guess I'm so depressed that I'm not even trying the obvious. Though the newsgroups mention it as a drive failure, so I was put off from the start. But, I'll check in a moment and post a second message.

Also does the drive spin up and is it detected by the bios?

Yes. It is the boot drive. It just won't mount.

Re:Welp

[Chacham]

# e2fsck -fv /dev/hda1
e2fsck 1.27 (8-Mar-2002)
e2fsck: Attempt to read block from filesystem resulted in short read while tryin
g to open /dev/hda1
Could this be a zero-length partition?
# fdisk /dev/hda

The number of cylinders for this disk is set to 9964.
There is nothing wrong with that, but this is alrger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versio0ns of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/hda: 255 heads, 63 sectors, 9964 cylinders
Units = cylinders of 106065 * 512 bytes

Device Boot Start End Blocks Id System
/dev/hda1 1 9964 80035798+ 83 Linux

Command (m for help): q

#

Re:Welp

[Chacham]

Even more.

Maxtor PowerDiag 3.04 shows the drive to have failed the SMART self test. When I ran the advanced test, it gave me a warning. But it said that PowerDPS can attempt to repair the errors. However, give the red sign and the yellow flashing, and the precautions it mentions, I assume that this is a bad thing. Who know? Maybe it'll erase all data on the drive. Therefore, I don't want to do it without knowing exactly what it plans to do.

I guess I'll call Maxtor tomorrow. Maybe they can help me here. Although, at the same time, I am afraid to rely on techinical support.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:Welp

[Chacham]

Just called Maxtor. They didn't know what PowerDPS was. He told me the error code just meant that something was changes in the BIOS or the OS. When I told him that it wasn't, and that the system has been fine for a while, he asked me about viruses. I said something like, "Linux doesn;t really have viruses." He replied to the effect of, "OK, I'll take your word on that..."

So, as expected, useless. They don;t even know what their own diagnostic software try to do. :(

Now its time to install Linux on another drive, and then back it up, and low level format it. If I can't mount it even off the other drive, I'll try another controller. I wonder if I can ghost it.

Re:Welp

[Chacham]

-b Doh!

I did (something like) a e2fsck -vfpb 32768 and it worked! I first tried a bad block test and got errors. So, I canceled that, and tried a standard filesystem check. That worked. So, there probably is failure on the drive, but I'm in business.

I took the other hard drive and installed a prety simple Debian distro on it. I am currently backing up the other hard drive. With nearly eighty gig to copy, it may take a few minutes. I hope to then plug in the old hard drive and see if it'll boot anyway.

I wonder if I can low level format it, and then just copy the old HD data back. I understand that I'll need to install the MBR, and run something to get the System.map up to date.

Re:Welp

[Chacham]

Finally. I'm back up! Woohoo! :-)

After copying everything, I reran the diagnostics software. It showed no errors, even on the advanced test. Strange, isn't it?

So, I swapped cables and turned it back into the primary master, and it booted without error.

Re:

[account_deleted]

Comment removed based on user account deletion

Re:Welp

[duren686]

I think the drive is pretty fscked already (badabum-PSSH)

Drives

[On Lawn]

At work, we've had one IBM and one Maxtor bite the bullet. Both were on (different) computers that wound up to have overheating DMA chips.