Centos recovering from Kernel panic on startup

[David Lloyd]

Ian,

> EXT3-fs INFO: recovery required on readonly filesystem.
> EXT3-fs: white acess will be enabled during recovery.

This probably means that the root filesystem itself is actually corrupt 
in some form or another. I assume that "white" means "write".

> hda: dma_intr: status=0x51 {DriveReady SeekComplete Error}
> hda: dma_intr: error0x40 {UncorrectableError }, LBAsect=221200, 
> sector=221205
> ide: failed opcode was: unknown
> end_request: I/O error, dec hda, sector 221205

That sounds like a hard disk error, i.e. hardware. That's certainly no 
good at all.

> JBD: IO error reading journal superblock

Some (most?) filesystems have special blocks called "superblocks". These 
contain very important information about the filesystem layout, where 
things are and such.

Generally, most filesystems keep a copy or two about the place just in 
case one gets corrupt. I think the journal driver is saying that it 
can't find the block which allows it to read the journal..

> EXT-fs: error loading journal,

...which causes that error.

> mount: error 22 mounting ext3
> mount: error 2 mounting none
> Switching to new root
> Swichroot: mount failed 22
> umount /initrd/dev failed: 2
> Kernel panic - not syncing: Attempted to kill init!

I'd suggest:

  1. Get a backup of the actual bits on the drive if you can
  2. Attempt to mount the driver WITHOUT the journal (it sounds like it's
     ext3; you might be able to mount it as ext2)

     BIG WARNING: Doing this step _may_ / _will_ lose data. Please do NOT
     complain, blame, sue or otherwise tell me I didn't tell or warn you.

Realistically, I'd get a rescue disk or an install disc with a 
sufficiently good set of tools (Gentoo is good, so are the modern 
Debian's) and quickly get all the important information off the drive.

Then see if it's still under warranty because it's probably going to get 
more bad blocks and worse.

DSL