Re: [Jfs-discussion] FS corrupt after kernel crash

[Dave K. <sh...@li...>]

On Sun, 2008-09-14 at 06:26 +0200, Geronimo Ma. Hernandez wrote:
> Hello,
> 
> after a system crash on a linux-box, the jfs-disk failed on fsck, so I got the 
> freshest jfs-util and did a manual fsck:
> -----------------------
> $> fsck.jfs -nv --omit_journal_replay /dev/hdd1
> 
> fsck.jfs version 1.1.13, 17-Jul-2008
> processing started: 9/13/2008 20.6.3
> The current device is:  /dev/hdd1
> Open(...READONLY...) returned rc = 0
> Primary superblock is valid.
> The type of file system for the device is JFS.
> Block size in bytes:  4096
> Filesystem size in blocks:  73258400
> **Phase 1 - Check Blocks, Files/Directories, and  Directory Entries
> ujfs_rw_diskblocks: read 0 of 16384 bytes at offset 73014460416
> Unrecoverable error reading M from /dev/hdd1.  CANNOT CONTINUE.
> Fatal error (-10015,30) accessing the filesystem (1,73014460416,16384,0).
> processing terminated:  9/13/2008 20:06:19  with return code: -10015  exit 
> code: 8.
> -----------------------
> 
> The disk is a 300Gb drive having just one partition (entire space).
> Kernel was Linux version 2.6.23.9
> The fs had about 70% of space used.
> 
> Any hint, how I could recover some/most data?

I don't have time right now to dig into whether or not I can hack fsck
to get past this error.  (It often gives up too easily.)  The quick
thing to try is to mount the file system read-only.  "mount
-oro /mountpoint"

If that doesn't work, I'll try to take a look at where fsck is failing
and see if I can patch it to get to continue with what it still can
recover.

> kind regards
> 
> Geronimo

Shaggy
-- 
David Kleikamp
IBM Linux Technology Center