Saving an injured IDE HardDrive

Beaud · 01-16-2008, 10:58 PM

Hi, I installed FC6 on my box, though it was the best idea to restore data from a ~80gb IDE HardDrive. I ran badblocks with nearly 100% badblocks after 10mins, canceled. I'm currently running gpart and I'm having read errors EIO near sectors(...) every 64 "sectors". As an additionnal information, I tried to mount the hard drive from the FC Rescue Mode. No result. fdisk can't find any valid partition. What's left to do now?
I guess I could afford losing some data if there is no choice but that.

Regards, Beaud.

Simon Bridge · 01-16-2008, 11:11 PM

You could frame it as art?

Technically you could just bitdump the contents and run an eyeball along it... got a free month or six?

Naturally, this is where you are glad of your rigorous and methodical backup policy.

kromberg · 01-17-2008, 08:02 AM

The term 'SOL' is coming to mind.

Keith

Simon Bridge · 01-17-2008, 10:24 AM

Roughly third to half of the official expansions of the abbreviation are relevant - yes. The one you mean is not listed, but I think we get the gist.
http://acronyms.thefreedictionary.com/SOL
SOL Sadly Outta Luck (polite form)
SOL Satellite Of Love
SOL Seafarer's Open Learning System
SOL Service Opportunities in Leadership
SOL Shazzam - Outta Luck (Gomer Pyle polite form)
SOL Shoot-Out Loss
SOL Short on Luck
SOL Simply Out of Luck (polite form)
SOL Sobbing Out Loud
SOL Society for Organizational Learning
SOL Somewhat Out of Luck
SOL Sorta Outta Luck
SOL Space-Occupying Lesion (medical)
SOL Standards Of Learning
SOL Statement of Liability
SOL Stuff Outta Luck (polite form)
SOL System Off Line

Emerson · 01-17-2008, 10:58 AM

http://www.acronymattic.com/results.aspx?q=SOL

tredegar · 01-17-2008, 12:04 PM

Beaud,
Looks like that disk is bad.
Very bad.
Sorry.

There's a linux utility called photorec that may be able to recover something. Worth a try?
Please google for it.

& Welcome to LQ

!

Beaud · 01-17-2008, 12:15 PM

Code:

Fedora Core release 6 (Zod)
Kernel 2.6.12-1.2798.fc6 on an i686

localhost login: hdd: task_in_intr: status=0x59 { DriveReady SeekComplete DataRequest Error}
hdd: task_in_intr: error=0x01 { AddrMarkNotFound }, LBAsect=1600866520, sector=160086520
ide: failed opcode was: unknown
hdd: task_in_intr: status=0x59 { DriveReady SeekComplete DatRequest Error }
hdd: task_in_intr: error=0x01 { AddrMarkNotFound }, LBAsect=160086520, sector=160086520
ide: failed opcode was: unknown
hdd: task_in_intr: status=0x59 { DriveReady SeekComplete DatRequest Error }
hdd: task_in_intr: error=0x40 { UncorrectableError }, LBAsect=160086520, sector=160086520
ide: failed opcode was: unknown
end_request: I/O, dev hdd, sector 160086520
Buffer I/O error on device hdd, logical block 20010815

I got that error from fdisk as well.

fdisk output:

Code:

Disk /dev/hdd: 81.9 GB, 81964302336 bytes
16 heads, 63 sectors/track, 158816 cylinders
Units = cylinders of 1008 * 512 = 516096

As for the table, it is empty.

Code:

160086527 unallocated sectors

Out of luck at mounting:

Code:

[root@localhost ~]# mount -t ntfs /dev/hdd /mnt
mount: unknown filesystem type 'ntfs'
[root@localhost ~]# mount -t vfat /dev/hdd /mnt
FAT: bogus number of reserved sectors
mount: wrong fs type, bad option, bad superblock /dev/hdd
       missing codepage or other error
       In some cases useful into is found in syslog...

Nothing useful using dmesg...

Code:

[root@localhost ~]# mount /dev/hdd /mnt
hfs: unable to find HFS+ superblock
mount: you must specify the filesystem type

In Rescue Mode...

Code:

Running anaconda, the Fedora Core rescue mode - please wait...
ERROR: asr: reading /dev/hdd[Input/output error]
ERROR: ddf1: reading /dev/hdd[Input/output error]
ERROR: hptx45: reading /dev/hdd[Input/output error]
ERROR: isw: reading /dev/hdd[Input/output error]
ERROR: jmicron: reading /dev/hdd[Input/output error]
ERROR: lsi: reading /dev/hdd[Input/output error]
ERROR: nvidia: reading /dev/hdd[Input/output error]
ERROR: pdc: reading /dev/hdd[Input/output error]
ERROR: pdc: reading /dev/hdd[Input/output error]
ERROR: sil: reading /dev/hdd[Input/output error]

Might be missing an error or two here..

Now badblocks:

Code:

sh-3.1# badblocks -v /dev/hdd
Checking blocks 0 to 80043264
Checking for bad blocks (read-only test): 256        256/    80043264
304        304/        80043264
305        305/        80043264
306        306/        80043264
307        307/        80043264
308        308/        80043264
...

Last but not least, gpart:

Code:

Warning: read error (EIO) near sector (...), skipping...
...

Thats about all I can see on my monitors for hours..

Roll your eyeballs on that.

Beaud · 01-17-2008, 12:18 PM

Thanks tredegar, appreciated, I'll look into it, otherwise I will attempt Spinrite, need to make sure the guy won't blame me for losing his data, even if its already lost.

tredegar · 01-17-2008, 12:23 PM

As I said, it looks bad.
But you cannot mount disks like this:

Quote:

Code:

mount -t ntfs /dev/hdd /mnt

You need to mount a partition to a directory, more like this:

Code:

mount -t ntfs /dev/hdd0 /mnt/baddisk

The directory baddisk must exist before you try the mount.
But I think the disk is bad, so this is unlikely to work

Give photorec a try, it is not just for photos (jpg's), but loads of other formats as well.

Beaud · 01-17-2008, 12:27 PM

Yeah, I've googled it, looks good. I'll give it a try right now. Thanks.

Emerson · 01-17-2008, 12:30 PM

Quote:

[root@localhost ~]# mount -t ntfs /dev/hdd /mnt
mount: unknown filesystem type 'ntfs'

NTFS support missing?

Beaud · 01-17-2008, 12:37 PM

Yes, it is missing... for some reason or maybe I am missing a lib.

Emerson · 01-17-2008, 12:41 PM

Not sure but I think FC6 didn't have NTFS support out of box.

tredegar · 01-17-2008, 12:57 PM

Quote:

Yes, it is missing... for some reason or maybe I am missing a lib.

You are missing a module:

Code:

tg@vaio:~$ lsmod | grep ntfs
ntfs                  103536  2
tg@vaio:~$

So, try, as root, modprobe ntfs

Beaud · 01-17-2008, 01:18 PM

FATAL, not found, I'll install it. With yum I guess