Messages in this thread | | | Date | Wed, 17 Dec 1997 09:23:26 -0500 (EST) | From | "Brian H. Trammell" <> | Subject | ext2_free_inode errors |
| |
Greetings,
Haven't been lurking dutifully recently, but even so, I don't remember any reports about the following error.
I'm running RH4.2 upgraded to 2.0.32 on a mail, web, dns, and ftp server. I'm getting a lot of the following in my /var/log/messages file:
Dec 14 01:02:23 discworld kernel: EXT2-fs warning (device 03:03): ext2_free_inode: bit already cleared for inode 53065 Dec 15 01:02:24 discworld kernel: EXT2-fs warning (device 03:03): ext2_free_inode: bit already cleared for inode 53065 Dec 16 01:02:22 discworld kernel: EXT2-fs warning (device 03:03): ext2_free_inode: bit already cleared for inode 53065 Dec 17 01:02:23 discworld kernel: EXT2-fs warning (device 03:03): ext2_free_inode: bit already cleared for inode 53065 Dec 17 09:11:02 discworld kernel: EXT2-fs warning (device 03:03): ext2_free_inode: bit already cleared for inode 53065
This appears to coincide with the redhat daily cron stuff at 1am. (The 09:11 access today was apparently related to my ls -aliR'ing the partition on /dev/hda3)
inode 53065 is /ankh/ftp/pub/rh5-i386/misc/src/install/libfdisk: total 470 53065 ?--------- 0 root root 0 Dec 31 1969 CVS
(this is the full RH5 image pulled down Dec 01)
I was able to remove this whatever successfully.
I also have a lot of
Nov 19 10:19:20 discworld kernel: EXT2-fs warning (device 03:04): ext2_unlink: Deleting nonexistent file (126508), 0 Nov 19 10:19:20 discworld kernel: EXT2-fs warning (device 03:04): ext2_free_inode: bit already cleared for inode 126508 Nov 19 10:19:21 discworld kernel: EXT2-fs warning (device 03:04): ext2_free_inode: bit already cleared for inode 126506 Nov 19 10:19:21 discworld kernel: EXT2-fs warning (device 03:04): ext2_unlink: Deleting nonexistent file (126506), 0 Nov 19 10:19:21 discworld kernel: EXT2-fs warning (device 03:04): ext2_free_inode: bit already cleared for inode 126506
but I can't find any inodes within that numeric range any longer on /dev/hda4.
Questions: Is this a known problem? Is it a serious problem? (i.e. if I'm seeing this problem on a production server, should I downgrade to an earlier kernel with an entirely different set of bugs? :) Is there anything more I can do to help its resolution?
TIA,
brian
P.S. still crunch time at work, so please cc me on any replies to the list. thx.
-------------------------------------------------------------------------- brian h. trammell <brian@altara.org> http://www.altara.org/~brian georgia tech civil engineering system administration "The problem with the rat race is that the winner is still a rat." 0xf00fc7c8
| |