Date: Tue, 27 Jul 2004 17:52:25 +1000 From: Peter Jeremy <PeterJeremy@optushome.com.au> To: Charles Sprickman <spork@fasttrackmonkey.com> Cc: freebsd-hackers@freebsd.org Subject: Re: disk recovery help Message-ID: <20040727075225.GL3001@cirb503493.alcatel.com.au> In-Reply-To: <20040726142446.P28049@toad.nat.fasttrackmonkey.com> References: <20040719191408.V28049@toad.nat.fasttrackmonkey.com> <20040720021432.O28049@toad.nat.fasttrackmonkey.com> <20040720092848.GD3001@cirb503493.alcatel.com.au> <20040720135157.Q28049@toad.nat.fasttrackmonkey.com> <20040722075723.GE3001@cirb503493.alcatel.com.au> <20040726142446.P28049@toad.nat.fasttrackmonkey.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 2004-Jul-26 14:30:08 -0400, Charles Sprickman wrote:
>I did get confirmation from Adaptec that it does go from the outside
>sectors in.
In which case one of the first things to be over-written would have
been the first superblock and fsck should complain and stop immediately
if this is invalid.
>I ended up using sysutils/ffsrecov to grab the alternate superblocks.
Maybe that was the tool I was thinking of.
>I have a reasonably OK fsck'd filesystem mounted now. I have another copy
>to work on, and my question there is this: When you run fsck it creates a
>"lost+found" directory to put files that are unreferenced anywhere (I
>think that's the terminology). At some point during the fsck, it starts
>spitting errors about there not being enough space in "lost+found". Is
>there any way to remedy that problem? Is there some way to "grow" the
>filesystem *before* fsck-ing it?
fsck will grow the lost+found directory if necessary but it can only
grow it to the limit of the direct blocks (12 filesystem blocks). The
only way to pre-grow the lost+found directory would be to mount the
filesystem read/write, create a large number of (preferably large)
filenames and then delete them[1]. I'm not sure if the fsck code can
understand indirect blocks in the lost+found directory so I don't know
if this would work. It also relies on the filesystem to be sane enough
for normal block allocation to work.
One option might be to force a read/write mount, rename lost+found,
umount the filesystem and redo the fsck until it fills the new
lost+found. Renaming a directory entry is fairly safe, especially if
you don't change the entry size. Alternatively, do this with fsdb -
which is definitely safe.
[1] Directories are only shrunk when an entry is created, not when one
is deleted.
--
Peter Jeremy
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040727075225.GL3001>