OK, so the next possibility is the fact that xfs is setting the block
size of the device to 512 bytes and that pushes some calculations over.
I am experimenting with some new code in 2.4.13 which removes this
chunk of code, the next thing to try may be this, I am running md
tests on it this morning and so far it looks good, but I do not
have access to enough disk to test for the size overflow.
I hope to have a 2.4.13 xfs patch out today, maybe we will have to
get you to try it.
Thanks for your patience on this one.
> Steve Lord wrote:
> > Yes, but the detection process may need some work - ext2 will probably
> > have radically different behavior on an error. It may throw up on the
> > filesystem, or you may have to run fsck afterwards to see what happened
> > to the fs.
> Hmm, using mke2fs 1.25 (with m set to 1), mongo runs a complete
> benchmark over the 1.2TB:
> Filesystem 1k-blocks Used Available Use% Mounted on
> /dev/md0 1135510828 20 1123788584 1% /local
> After unmounting, I ran e2fsck:e2fsck 1.25 (20-Sep-2001)
> /dev/md0: clean, 11/293076992 files, 9177898/293055600 blocks
> Same steps as above with the current version of e2fsprogs on a RedHat
> 7.1 system showed no different result (e2fsck 1.23, 15-Aug-2001 for EXT2
> FS 0.5b, 95/08/09)