Hi all,
I'm running ext3 with Linux 2.4.19pre5aa2 (and thus ext3 0.9.17)
and e2fsprogs 1.27 on an x86 laptop with an IDE disk (with DMA
enabled). The partition in question was mounted with the
data=journal and nosuid options set. All the partitions are set to
be fscked on bootup every 30 days (although I don't think this
is strictely necessary, better safe than sorry).
ext3 has been working fine for a number of months. Earlier
today, I rebooted the system (AFAIK, the partition in question
unmounted cleanly, although I wasn't paying attention to it
particularly); when linux started to load again, it chose to
do a periodic fsck of all the partitions. Everything was clean
except for /dev/hda5, which had many fsck warnings like so:
Inode 8387: i_blocks is 2, should be 0. FIXED
(sorry, don't have the exact text, but that was the gist of it)
Those errors occurred for a lot of inodes (70+), a lot of them
numbered sequentially (8387->8393, 14424->1453, etc). For most
of these inodes, i_blocks was 2 and should have been 0 -- other
variations included i_blocks at 36 when it should have been 34,
and 76 instead of 2. According to fsck, all the problems that
it detected were fixed.
As far as I can remember, the only unusual activity that
happened on that partition is that it ran out of disk space
several times. Could that be the problem?
Please let me know if you need any more information, or if
you have any idea on what the problem might be.
Cheers,
Neil
--
Neil Conway <neilconway@rogers.com>
PGP Key ID: DB3C29FC