What steps will reproduce the problem? 1. shut down EeeDora or reboot EeeDora 2. on next boot an fsck is forced as the / filesystem wasn't cleanly unmounted 3.

What is the expected output? What do you see instead? Well - No forced fsck...

What version of the product are you using? On what operating system? 2008-01-25

Please provide any additional information below.

Same error, except in my case it often corrupts the system so badly (cloned/crosslinked inodes) that the whole install is borked. If anyone figures this out, please let us know. Eeedora is perfect until I reboot.

Is this is for an install on the Internal Drive /dev/sda1 ?

Do you get the same fsck problem if you (a) boot the Eee, (b) immediately shut-down the Eee (without opening any files, etc, in between) ?

As far as borking the volume goes, the ext2 filesystem is less resilient than the more typical ext3 (but we need ext2 to avoid wearing out the flash drive).

Same for me, eeedora is installed on an external 8GB SD card mapped to /dev/sda in the BIOS (the internal SSD is unmodified). I have /dev/sda1 as / with no other mounts apart from a 520MB swap partition /dev/sda2.

The initial fsck always finds /dev/sda1 was not cleanly unmounted, despite the fact that an e2fsck at the end of /etc/init.d/halt (just before the final poweroff) claims it is clean.

This suggests it is some kind of driver/caching problem for the builtin SD reader. I've added

hdparm -r 1 /dev/sda

to /etc/init.d/halt after the root filesystem is remounted readonly and this seems to have stopped the problem so far. I don't whether it is actually safe to recommend that generally.

Ok, this may not be the answer I've just hit an fsck after about 5 on/off cycles, but it's no longer doing it every time.

My BIOS clock was going back 1 hour on each reboot, causing the last mount time to be in the future and triggering an e2fsck.