I'm getting this error after running xfs_repair on a disk:
fatal error -- bad magic # (0x0) for directory data block (bno 0 fsbno 7)
It all started with smartmontools telling there was a bad block, via
http://oss.sgi.com/archives/xfs/2005-08/msg00099.html I tried to determine what
file exactly it was, but an md5sum of it said everything was ok. I tried both
the manual calculation and the convert method to get the fsb, both giving me a
Giving no result, I used badblocks in non-destructive mode on an area around
the fsb I calculated, and it it seems it was after that the disk listed nothing
when mounted. Could badblocks in non-destructive mode have caused that?
Then, an umount/mount didn't help and an xfs_check and xfs_repair later it gave
that fatal error.
There's also a lot of other errors in that repair-log:
bad magic number 0xa371 on inode 148
bad version number 0x1c on inode 148
bad inode format in inode 148
bad magic number 0xfd35 on inode 149
bad version number 0xffffffb9 on inode 149
bad (negative) size -8929491557123885919 on inode 149
If wanted I can attach logs of the xfs_check/repair, smartctl error ...
Thanks for any help, please CC me as I'm not subscribed to the list.
Invite your mail contacts to join your friends list with Windows Live Spaces.
[[HTML alternate version deleted]]