>
> On an XFS note (unrelated to 2.4.10 or 11) - both my laptop and
> desktop systems (2.4.9, 2.4.11) ALWAYS say on bootup that recovery is
> required on the root filesystem - even though it's been shut down
> cleanly. My laptop's been doing this for months and I think the
> desktop box started doing it in the last week or so (running 2.4.7).
> There's nothing unusual about the boot or shutdowns (everything seems
> clean) except this one thing :-/
> RedHat7.1 with kgcc. I have cvs xfsprogs also.
This line in the /etc/init.d/halt script could be the culprit:
#echo $"Remounting remaining filesystems (if any) readonly"
mount | awk '/ext2/ { print $3 }' | while read line; do
mount -n -o ro,remount $line
done
It is not xfs friendly.
Steve
>
> from dmesg on every boot:
> ...
> XFS mounting filesystem ide0(3,1)
> XFS: WARNING: recovery required on readonly filesystem.
> XFS: write access will be enabled during mount.
> Starting XFS recovery on filesystem: ide0(3,1) (dev: 3/1)
> Ending XFS recovery on filesystem: ide0(3,1) (dev: 3/1)
> VFS: Mounted root (xfs filesystem) readonly.
> ...
>
> I could probably xfs_repair (or just mount) the desktop disk with
> a rescue CD to fix it, but the laptop hasn't enough memory to run the
> rescue CD and I'd have to pull the @#$@#$@ disk out and move it to
> another machine to sort it out - big hassle.
>
> Any ideas?
> Maybe it's corruption left over from an earlier kernel version that's
> never being fixed?
> A friend of mine found it also and thought it might be 'cos we upgraded
> from earlier RH/XFS versions rather then re-installing.
>
> cheers,
> robin
>
> (*) allocates memory (eg. 400M when I have 384M of ram) and touches
> many of the pages whilst allocating, and then touches random pages
> 10k times once the full 400M is reached. It's designed to simulate an
> app that has bad data locality.
|