fsck: could this be a zero-length partition?

Hello,

I'm afraid I'm a complete newbie; I'm posting out of desperation.

I have Debian on one partition of a Dell Inspiron 5100 Laptop. I think the Kernel is 2.6.17 or so. The other partition has Windows XP. For the past few weeks, Linux has been acting a bit too much like Windows, i.e. randomly freezing so I had to shut it down. During this time, windows stopped working altogether.

Finally, today, while trying to start my computer I got this message: GRUB loading . . .
Error 18. After restarting, and going into the boot menu, then exiting the boot menu, I got another message: Primary hard disk not found.

I was able to boot from a Knoppix 4 cd. I can mount my Windows partition (hdc2), but unfortunately not my Linux partition (hdc3).

I used to get the message: Could not mount device. The reported error was: mount:wrong fs type, bad option, bad superblock on /dev/hdc3. Missing codepage or other error. Now, excitingly, I get a different message when I try to mount the hdc3 partition: mount: I could not determine the filesystem type, and none was specified.

So next I tried fsck /dev/hdc3. The first time, I got a lot of messages saying, "Attempt to read block [insert number here] resulted in short read. Ignore error?[y?]" To which I would hit y. Then I would get something like "Block [insert number] should be 104. Force rewrite?[y?]" To which I would hit y. (I'm sorry, I don't remember exactly what it asked, and it won't repeat it). I did this a few hundred times, and then decided that it probably wasn't working, so I tried "n" instead of "y" and it aborted.

Now when I try fsck /dev/hdc3 I get fsck.ext2:
Attempt to read block from filesystem resulted in short read while trying to open /dev/hdc3
Could this be a zero-length partition?

Next I tried reiserfsck /dev/hdc. I got a message saying:

bread: Cannot read the block (2): (Input/output error).
Aborted

It comes with a nice paragraph telling me that it's probably a hardware problem and I should get a new hard drive.

I've also tried sfdisk -l /dev/hdc. I got this:

Disk /dev/hdc: 58140 cylinders, 16 heads, 63 sectors/track
read: Input/output error

sfdisk: read error on /dev/hdc - cannot read sector 0
/dev/hdc: unrecognized partition table type
No partitions found

So its all looking pretty dire for my hard drive. The question is: is there ANY way to get my data off of the hdc3 partition or is it gone forever?

Remember, I'm a newbie, so no big words or complicated terms, and if you need more information you'll have to tell me where to get it and how. And yes, I have Googled this problem, but I wasn't able to understand many of the solutions.

Thank you for your help. I can't tell you how much it would mean to me to get my data back.

zlya

First double check whether you mean /dev/hda3 instead. After booting with a live distro, run "fdisk -l" to list all the drives and partitions.

It is possible that the first superblock is damaged. Try as root: "dumpe2fs /dev/hdc3"if this filesystem is ext2 or ext3. You can mount using a backup superblock. Add "bs=8193" to the mount command as one of the options. You might also try adding "error=continue".

There is a Steve Gibson's commercial program called Spinrite that might help recover your drive.
http://spinrite.info/

Thank you for your help. I was finally able to mount the drive and recover my data. In my desperation I was trying every Knoppix tool that looked remotely hard-disk related, and I noticed that QTParted thought that hdc3 was an ext2 partition, while /etc/fstab thought it was an ext3 partition. So I the entry in /etc/fstab to ext2, and I was able to mount the drive. Not the most orthodox of approaches, I'm sure. Now the computer still won't start, but at least I've got my data back.

Thanks again,

zlya