2.2.1 ext2-related crash

Olaf Titz (olaf@bigred.inka.de)
Thu, 11 Feb 1999 12:03:04 +0100


Linux 2.2.1, K6-233, NCR815, no signs of trouble with the disk (IBM
DCAS) or memory so far.

Feb 11 06:42:05 bigred kernel: EXT2-fs warning (device 08:01): ext2_free_inode:
bit already cleared for inode 3836
Feb 11 06:44:22 bigred kernel: attempt to access beyond end of device
Feb 11 06:44:22 bigred kernel: 08:01: rw=0, want=167772158, limit=41509
Feb 11 06:44:22 bigred kernel: EXT2-fs error (device 08:01): ext2_readdir: directory #4025 contains a hole at offset 0
Feb 11 06:44:22 bigred kernel: attempt to access beyond end of device
Feb 11 06:44:22 bigred kernel: 08:01: rw=0, want=473270776, limit=41509

Then the system crashed and was rebooted (perhaps by softdog). The
partition on 08:01 showed severe damage (lots of "entry has
deleted/invalid inode" type errors, luckily mostly for /dev entries).

Other system logs tell the type of activity on the filesystem:

Feb 11 06:44:14 bigred su[4635]: SU: 'su nobody -c find / [lots of stuff]'
that's Debian's checksecurity, which scans the whole filesystem.
So I assume the problem is related to high load on the FS.
Also this:
Feb 11 06:44:19 bigred squid[193]: storeDirWriteCleanLogs: Starting...
The squid logfiles were also damaged (as the only files on a partition
other than root).

Olaf

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/