[TriLUG] data recovery - a modest how-to

[Brent Verner]

Hi,

  I recently had an IBM-DTLA-307045 die on my box at work, and sure
enough, I'd not backed up some (more than I'd care to lose) work
in progress... panic.

  The partition would not mount, nor fsck, due to read errors on the
(primary) superblock, so I forcibly pulled everything off the partition
that I could get, just incase I'd have to do a manual recovery ;-)

  # dd conv=noerror if=/dev/hdc7 of=hdc7

Let this guy run as long as it needs.  Next, try to see where a
backup superblock is.  We can use this backup superblock instead of
the primary superblock, since that one is /known/ to be inaccessible.

  # mke2fs -n /dev/hdc7
  mke2fs 1.25 (20-Sep-2001)
  Filesystem label=
  OS type: Linux
  Block size=4096 (log=2)
  Fragment size=4096 (log=2)
  152960 inodes, 305605 blocks
  15280 blocks (5.00%) reserved for the super user
  First data block=0
  10 block groups
  32768 blocks per group, 32768 fragments per group
  15296 inodes per group
  Superblock backups stored on blocks: 
    32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208

Ok, that last line is the location of backup superblocks, walk the 
list until we find a good one.

  # e2fsck -b 32768 /dev/hdc7

This rebuilt the filesystem so it could be mounted again and copy off
the data.  Only one file (~/.bash_history) lost -- woohoo!!!

Remember kids, backup.  But incase you don't, hopefully this might
help someone out of the mess I found myself in :-)

Oh yeah, if you happen to have one of the IBM-DTLA-307045 drives, 
replace it today.  A quick google search will reveal horrors of this
model.  FWIW, I replaced the drive with one of IBM's IC35L040AVER07-0
drives, which I surely hope isn't afflicted with whatever was causing
the DTLA line to fail.

cheers.
  Brent

-- 
"Develop your talent, man, and leave the world something. Records are 
really gifts from people. To think that an artist would love you enough
to share his music with anyone is a beautiful thing."  -- Duane Allman