Re: EXT2 directory corrupted
You have a hosed root directory and/or inode. You should say 'no'
to these questions and see if that gets you past the problem.
Barring that there is debugfs that I once used back in the 2.0
kernel days to fix an obnoxious problem like this, turned out the
corruption was very minor, but the fsck program was completely
confused by it. These programs often have difficulty with problems
in the root directory because of assumptions in the code.
Mount the partition in RO, and check the inode for /lost+found:
# ls -lid /lost+found
11 drwxr-xr-x 2 root root 16384 Feb 20 21:18
If yours is also 11 then answer the question about reconnecting it
with 'n'. Unless you're storing important files there ~:^) You
will be able to recover the space later in the fsck process. If
there are any other unconnected inodes, do not attempt to have them
reconnected either. I would recommend running fsck again after it
is done. You will have to do an mklost+found after all this as that
directory will be gone.
Christian Jaeger wrote:
> At 11:07 Uhr -0700 5.7.2001, Peter Canning wrote:
> >I've recently encountered some strange file system corruption problems.
> >Here's the results of running e2fsck
> ># e2fsck -f -b 40961 /dev/sdb6
> >e2fsck 1.21-WIP, 14-Jun-2001 for EXT2 FS 0.5b, 95/08/09
> >Pass 1: Checking inodes, blocks, and sizes
> >Pass 2: Checking directory structure
> >Pass 3: Checking directory connectivity
> >Unconnected directory inode 11 (???)
> The problem probably is, that on most of my ext2 filesystems
> lost+found *is inode 11*. That means that e2fsck is probably trying
> to move lost+found to itself.
> However I have also ext2 filesystems with lost+found being other
> inode numbers. So maybe the solution is to say NO to the above
> question from fsck, then manually removing lost+found, creating a new
> one, and fscking again. Of course I don't take responsibility for
> To UNSUBSCRIBE, email to email@example.com
> with a subject of "unsubscribe". Trouble? Contact firstname.lastname@example.org