On Fri, Oct 12, 2001 at 03:14:32PM +1100, Nathan Scott wrote:
> hi,
Hello,
> > It doesn't fix me. I'm still broken.
>
> That makes some sense, at least (i.e. that switching off quota
> doesn't fix the problem).
Switching off quota did not fix my problem. It did work for Derek though.
Oddness.
> > I'm willing to give anything a try at this point... The box in question
>
> I would suggest trying the CVS code if you can - that is now at
> 2.4.13-something and there's plenty of changes in there. Maybe
> this problem has been diagnosed & fixed in the base kernel - the
> fact that earlier versions work OK suggests (but doesn't prove)
> that the problem is coming from outside XFS.
That fixes it. The box is now running 2.4.13-pre1-xfs with no problems,
and Quota (and XFS Quota) are both enabled.
Another oddity, the "other" box is running 2.4.10 with an identical
40G lvm/XFS filesystem and is experiencing no problems at all. I'm still
unable to create new lvm volumes and mkfs.xfs them, but the existing
filesystem built under 2.4.7 is working flawlessly. I guess the question
is: "should I trust it"? Or should I rsync the filesystem now to the
2.4.13-pre1 box to be safe? (remember, identical boxes)
> > Oddly enough, the "rescue" 2.4.7-xfs kernel that they were built with work
> > fine... one box was able to create and format a large 40G xfs volume
> > yesterday with the "rescue" 2.4.7-xfs boot kernel, and I've been fighting
> > with the other machine to do the same under the "normal" 2.4.10-xfs boot
> > kernel. No dice.
> >
>
> I'd suggest booting the 2.4.7-xfs kernel on both machines for now,
> until this issue is resolved.
As these machines aren't really "production", but more of a personal
playground, I'd actually prefer to run the latest CVS snapshot.
> cheers.
No worries :)
- Ian C. Blenke <ian@xxxxxxxxxx> <icblenke@xxxxxxx>
|