Looking closer, it looks like the ext3 superblock is still intact but my data is hosed. I'll start from scratch and recover what I can from backups...after running badblocks on all the drives for that ultra-paranoid level of fun. T 2008/5/1 Twigathy <twigathy@xxxxxxxxx>: > Hi, > > I'm in a bit of a fix here - I was expanding an mdadm raid5 array from > 5 to 6 disks. During the expand, I had to shut down the machine to do > some fiddling about with its guts, retiring an old 4x200GB raid5 > actually. > > When I booted the machine back up again the array did not recover and > carry on reshaping. I tried various different tricks to see if I could > force-mount it, and I fear I may have wiped the array for good here... > > The position I'm in now: > md5 is (was?) a 6x600GB raid5, being expanded from 5 disks to 6. > The UUIDs on 5 of the disks match up > The UUID on the other disk is different (uhoh) > The disk with a different UUID has this info in --examine: > Reshape pos'n : 443168960 (422.64 GiB 453.81 GB) > Delta Devices : 1 (5->6) > so mdadm has been recording where the reshape got to... > > The other disks in the array do *not* have this reshape info in > them...I think this is probably because I force-assembled somewhere > down the line and screwed things up a bit. > > I can mount the 5-disk broken array and see about 1/4 of my data - > consistent with how far the reshape got before I shut down the > machine. > > What I would like to do is to assemble the array and continue > reshaping where it left off - at that 422.64GiB point. Where do I > start? Can I even do this? Is my data gone for good? :-( > > Thanks > > Tom C > -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
