Had a system crash the other day which fsck kindly fixed for me (or so I thought).
What started it was an unrecoverable write error. We decided to shutdown and run an fsck. This found a read error and offered to salvage it. The messages was
cannot read block 641322, i=41219 owner=bin, mode=40755, size 512, dir ? salvage y/n. We said yes and then got the following:-
missing '.' i=41219 node=40755 size=512 dir=/udk/usr/lib/locale/fr_FR.850/LC_MESSAGES fix y/n. We said yes. We then got:-
writing zeroed block 641322 to disk.
fsck then continues okay and we assumed that all was fixed. However, I find that we now have a directory which has a directory within it with the same inode! This is /udk/usr/lib/locale/fr_FR.850/LC_MESSAGES its inode is 41219 and within it there are two directories ..;28860 with inode 41218 and .;28860 with inode 41219! I believe these are spurious entries.
This causes problems with our full system backup (it loops around) Other scripts that scan the file system stop working once these get into this directory.
So anybody have an idea on how to fix this? We cannot delete it or fix it with fsck (I've tried -s -ofull -D).
Any suggestions would be appreciated. I've already had the reload SCO suggestion!
Mike
What started it was an unrecoverable write error. We decided to shutdown and run an fsck. This found a read error and offered to salvage it. The messages was
cannot read block 641322, i=41219 owner=bin, mode=40755, size 512, dir ? salvage y/n. We said yes and then got the following:-
missing '.' i=41219 node=40755 size=512 dir=/udk/usr/lib/locale/fr_FR.850/LC_MESSAGES fix y/n. We said yes. We then got:-
writing zeroed block 641322 to disk.
fsck then continues okay and we assumed that all was fixed. However, I find that we now have a directory which has a directory within it with the same inode! This is /udk/usr/lib/locale/fr_FR.850/LC_MESSAGES its inode is 41219 and within it there are two directories ..;28860 with inode 41218 and .;28860 with inode 41219! I believe these are spurious entries.
This causes problems with our full system backup (it loops around) Other scripts that scan the file system stop working once these get into this directory.
So anybody have an idea on how to fix this? We cannot delete it or fix it with fsck (I've tried -s -ofull -D).
Any suggestions would be appreciated. I've already had the reload SCO suggestion!
Mike