xfs
[Top] [All Lists]

Re: xfs_repair unable to repair a corrupt filesystem

To: "Russell G. Howe" <rhowe@xxxxxxxxxx>
Subject: Re: xfs_repair unable to repair a corrupt filesystem
From: Nathan Scott <nathans@xxxxxxx>
Date: Mon, 24 Feb 2003 16:27:59 +1100
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: <20030224043829.GA20677@lithium.sta.man.ac.uk>
References: <20030223081255.GA14077@lithium.sta.man.ac.uk> <20030223215931.GA798@frodo> <20030224033309.GA20260@lithium.sta.man.ac.uk> <20030224034815.GC798@frodo> <20030224043829.GA20677@lithium.sta.man.ac.uk>
Sender: linux-xfs-bounce@xxxxxxxxxxx
User-agent: Mutt/1.5.3i
On Mon, Feb 24, 2003 at 04:38:29AM +0000, Russell G. Howe wrote:
> On Mon, Feb 24, 2003 at 02:48:15PM +1100, Nathan Scott wrote:
> 
> http://rhowe.sfarc.net/xfs_repair_after_nblocks_fiddle.txt.gz
> 
> (and yes, I know xfs_repair says near the end that nblocks is 0, but
> xfs_db definately reported setting it to 189764)
> 
> As an example of how the filesystem behaves:
> 
> [rhowe@gonzo] ~ $
> 5:31AM
> Connection to gonzo.sfarc.net closed.
> [root@xiao] ~ # cd /var
> [root@xiao] /var # ls
> [kernel message]
> Filesystem "ide1(22,67)": corrupt dinode 2222, extent total = 2, nblocks
> = 0, Unmount and run xfs_repair
> [return to our regularly scheduled prompt]
> ls: Scarface.avi: Unknown error 990
> account  cache  local  log.nmbd    mail           run    www
> autofs   games  lock   log.smbd    oldlost+found  spool
> backups  lib    log    lost+found  rhowe          tmp
> [root@xiao] /var # ls rhowe
> ls: rhowe/Scarface.avi: Unknown error 990
> (this is accompanied by a similar kernel error to the one above, but
> with 6296608 as the dinode number.)
> 
> I don't need the /var/rhowe directory at all, as soon as I can fix this
> filesystem, I will be rm'ing it.
> 
> > > I don't need the file referenced by inode 2222, so losing it is of no
> > > consequence. I'd actually forgotten about its existence!
> > 
> > Is inode 6296608 a copy of this file?  (seem to have the same size
> > and both seem to be corrupted in the same way ...strange).
> 
> No idea. How could I find out?

Can you run the same (first) xfs_db command I sent to print
out the inode, except change the 2222 to 6296608?  I wonder
if these two inodes are pointing at the same extents...

> Maybe I was playing with hard links or
> something? (although I don't remember doing so...)

No, a hard link would be two distinct directory entries with
the same inode number - not quite the same as what we have
here.

cheers.

-- 
Nathan


<Prev in Thread] Current Thread [Next in Thread>