I'm running a server (Dual C533) under relatively high load with 2.2.14
patched with Ingo's md patch and an autodetected RAID1, mounted
root.  I've been noticing occational ext2fs errors like...

EXT2-fs warning (device md(9,0)): ext2_free_inode: bit already cleared for
inode 1206444
EXT2-fs warning (device md(9,0)): ext2_free_blocks: bit already cleared
for block 2432799
EXT2-fs warning (device md(9,0)): ext2_unlink: Deleting nonexistent file
(1206444), 0
EXT2-fs warning (device md(9,0)): ext2_unlink: Deleting nonexistent file
(1206439), 0


I'm not quite sure if this is the result of a failing disk, out of sync
disks due to failed writes, or bugs in either the raid or ext2 code.  I'm
less suspicious of the ext2 code than I am of the raid.  My main concern
is fixing the errors without munging the data.  I am most definately going
to switch to 2.2.17 with the ide+md patches but am also considering doing
one or more of the following:

1) Using raidhotremove on the second drive to get the system resync with
the spare disk.

2) Taking the system completely down and running a full chkraid.

3) or Just upgrade the kernel and let e2fsck handle the cleanup.


Any advice or info on past experiences with a similar situation would be
greatly appreciated.  Thanks,

-Kristian Hoffmann
<[EMAIL PROTECTED]>


-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to [EMAIL PROTECTED]

Reply via email to