I have the following error reported in my logs, I've had it for
quite
some time and through several reboots
EXT4-fs (md2): error count: 1
EXT4-fs (md2): initial error at 1452554514: ext4_lookup:1343: inode
1308162
EXT4-fs (md2): last error at 1452554514: ext4_lookup:1343: inode 1308162
md2 is my root partition, I ran "find / -mount -inum 1308162" find
didn't return any value.
How can I track down this error, identify what file it refers to and
clear the error out? if it matters, its an older Fedora system (16 to
be exact)
Thanks, Jeff
Thanks everybody for the suggestions, I figured running fsck would fix it,
I should have mentioned that in the past I've attempted to run fsck but it
hadn't cleared, probably how I ran it.
running fsck -fn /dev/md2 does in fact produce a handful of errors "inode
xxx ref count is 2 should be 1" plus a few other errors.
Since the machine is remote from me I'd like to issue a reboot and force
the disk to clean (I do have access to the system, just need to make a few
calls).
So my question is, after doing a little bit of searching can somebody
confirm this approach?
I thought (possibly incorrectly) I could force a fsck on reboot by
touching the file "/forcefsck" but searching further I think I will also
need to create the file "/etc/sysconfig/autofsck" and put these two lines
in:
AUTOFSCK_DEF_CHECK=yes
AUTOFSCK_OPT="-y"
hopefully this would clear the filesystem on a reload.
The other comment I saw would be to put a "-y" in the /forcefsck file
Am I on the right track? I'd prefer to avoid a drive over to the
datacenter and arranging access even though its only 15 minutes from home.
Booting the system on recovery media although not impossible will be
difficult.
Thanks, Jeff