Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 03 Nov 2004 01:26:59 -0500
From:      David Gilbert <dgilbert@dclg.ca>
To:        freebsd-current@freebsd.org
Subject:   More running-with-scissors and fsck.
Message-ID:  <41887A33.2060608@dclg.ca>

next in thread | raw e-mail | index | archive | help
Well... the recoverdisk is a cool tool.  I can see it coming in handy 
many times.  While there isn't much documentation, I've tended to use it 
by creating a vnode md disk of exactly the right size to recover the 
partition in question.

And I have a problem now.  recover disk finishes fine.  In fact, of the 
40G disk, it only can't recover about 210K.  That's a tiny fraction of 
1% unrecoverd.

However, it's all right around something key.  When I run fsck on the 
recovered image, fsck continuously complains that it can't create 
lost+found.

On the origional disk, of the 210K that's bad, most of it is near the 
start of the partition.  I suspect inode 1 and friends are not recovered 
... and thus all zeroed in the new md partition.

So... 99.99% of the data is there in good form.  How do I recover this 
... force a root to exist?

create an identical partion, newfs it and blindly copy the first 200 or 
so K?

Need someone who knows something about the structure I'm stomping on.

Dave.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?41887A33.2060608>