On Mon, Sep 24, 2012 at 6:12 PM, David Sterba [off-list ref] wrote:

On Mon, Sep 24, 2012 at 03:02:39PM +0100, Hugo Mills wrote:

quoted

   Out of interest, does mounting with -o recovery help at all? (I'm
not expecting it to do much if your chunk tree's gone, but it might do
something).

The -o recovery has access to the respective tree roots, but the
contents may be destroyed already. The chunk tree is not deep, I can see
height 1 on a 6 disk array (though lightly used, 1 node, 8 leaves) and 3
disk array (1/7 TB used, 1 node, 29 leaves). So it's quite a small
amount of data to destroy the chunktree completely, COW will lower the
chances a bit.

Yeah, the whole tree is gone, I'm pretty sure of it since the first
20-50GB has been wiped from the drive and the mentioned address is in
the beginning of that part. I just wonder if there is any chance of
the older versions of the chunk tree still being somewhere and how to
find them. I doubt it's an easy feat though.

Rebuilding from scratch does not look simple, the available information
is stored in BLOCK_GROUP_ITEMs or INODE_ITEMs and covers portions of the
chunks. Given that the device tree would be probably damaged as well,
the amount of information to do cross-check is not high. Maybe replaying
the chunk creation logic can save some guesswork.

Replaying chunk creation logic would not help that much, since the
drive has been resized a few times and had other operations that have
modified the chunk tree as well. The array itself is not that complex
(2 drives), but it's still not as simple as a single drive array.

Regards,
--
Sami Haahtinen
Bad Wolf Oy
+358443302775