[SWLUG] Data Recovery from unmountable disk

[Toby Maxwell-Lyte]

Hi Folks,

Discovered yesterday that the primary hard disk in one of our RH9 linux 
boxes has failed. It will no longer boot. Yes, I know I should have 
upgraded to fedora or some other disti, but it's been working fine for 
months...

Got the following in the logwatch email...

--------------------- Kernel Begin ------------------------ 

WARNING:  Kernel Errors Present
   EXT3-fs error (device ide0(3,2...:  1Time(s)
   end_request: I/O error, dev 03:02 (hda)...:  1Time(s)
   hda: dma_intr: error=0x40 { Uncorrect...:  1Time(s)
   hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }...: 
   1Time(s)

 ---------------------- Kernel End ------------------------- 

Oct 15 04:03:06 hda: dma_intr: status=0x51 { DriveReady SeekComplete
Error } Oct 15 04:03:06 hda: dma_intr: error=0x40 { UncorrectableError
}, LBAsect=1522600, sector=1310920 Oct 15 04:03:06 end_request: I/O
error, dev 03:02 (hda), sector 1310920 


So in my infinite wisdom I thought restarting it would be a good idea, I 
thought it would run efsk and sort everything out... that was clearly a 
mistake as it didn't boot after that.



I have booted it using a SystemRescueCd http://www.sysresccd.org

The disk is showing up as /dev/hda but none of the partitions appear...

% ls /dev/hd*
/dev/hda  /dev/hdb  /dev/hdb1  /dev/hdb2  /dev/hdb3  /dev/hdc

When I try to run fdisk -l /dev/hda there is no output... but a load of 
errors

Oct 17 11:01:20 [kernel] hda: task_no_data_intr: status=0x51 { 
DriveReady SeekComplete Error }
Oct 17 11:01:20 [kernel] hda: wcache flush failed!
Oct 17 11:01:21 [kernel] hda: task_in_intr: status=0x51 { DriveReady 
SeekComplete Error }
Oct 17 11:01:21 [kernel] ide0: reset: success
                - Last output repeated 16 times -
Oct 17 11:01:25 [kernel] hda: recal_intr: status=0x51 { DriveReady 
SeekComplete Error }
Oct 17 11:01:26 [kernel] ide0: reset: success
                - Last output repeated 126 times -
Oct 17 11:02:02 [kernel] hda: task_no_data_intr: status=0x51 { 
DriveReady SeekComplete Error }
Oct 17 11:02:02 [kernel] hda: task_no_data_intr: error=0x04 { 
DriveStatusError }
Oct 17 11:02:02 [kernel] hda: wcache flush failed!


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

% mke2fs -n /dev/hda
mke2fs 1.38 (30-Jun-2005)
/dev/hda is entire device, not just one partition!
Proceed anyway? (y,n) y
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
5020064 inodes, 10036941 blocks
501847 blocks (5.00%) reserved for the super user
First data block=0
307 block groups
32768 blocks per group, 32768 fragments per group
16352 inodes per group
Superblock backups stored on blocks:
        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 
2654208,
        4096000, 7962624

% e2fsck -b 32768 /dev/hda
e2fsck: Attempt to read block from filesystem resulted in short read 
while trying to open /dev/hda
Could this be a zero-length partition?

Can anyone help point me in the right direction? It seems that the 
partition table has disappeared... is there any way of getting it back?

Many thanks,
Toby

Web:   http://www.eazyweb.net
Email: Toby.Maxwell-Lyte at eazyweb.net
Tel:   +44 (0)29 20 344006
Fax:   +44 (0)29 20 344008
Skype: eazyweb_uk