Eric Sandeen wrote:
> Only thing I can think of is that this might be a >1TB problem?
> Trying to remember - have you done any tricks at mkfs time to keep inode
> numbers under 32 bits?
Lemme think. We talked about it, but in the end it wouldn't work with
r1.0.1 for some unknown reason and I upgraded to CVS TOT and then 'mkfs.xfs
/dev/md0' worked just fine.
> yocum@xxxxxxxx wrote:
> > OK, I'm using my kgcc compiled 2.4.8-xfs kernel, and I'm still seeing inode
> > corruption, even on a "static" filesystem (i.e., mounted noatime and no
> > writes occuring).
> > This is a very bad thing.
> > I'm about to punt and reformat the entire partition as XFS, and if it
> > happens again, then I'll have to reformat as ext2, which I really don't want
> > to do since it's 1.12TB.
> > If anyone has any ideas/patches please, please send 'em my way.
> > Thanks,
> > Dan
> Eric Sandeen XFS for Linux http://oss.sgi.com/projects/xfs
> sandeen@xxxxxxx SGI, Inc.
Sloan Digital Sky Survey, Fermilab 630.840.6509
SDSS. Mapping the Universe.