Change in ataraid (geometry?) between RELENG_5 and RELENG_6?

[Andrew Reilly]

On Fri, Sep 22, 2006 at 03:06:58PM +1000, Norberto Meijome wrote:
> On Fri, 22 Sep 2006 13:50:01 +1000
> Andrew Reilly <andrew-freebsd at areilly.bpc-users.org> wrote:
> 
> > I just tried to do an in-place upgrade from RELENG_5 to RELENG_6
> > on a system that was running an ataraid mirror on a pair of SATA
> > disks.  The root file system (256M) mounted fine, but fsck -p
> > fails (unable to find a superblock, from memory) under the new
> > kernel on my /usr partition, which is about 73G (most of the 80G
> > of the disks).  Falling back to the RELENG_5 /boot/kernel and
> > fsck finds no problem.
> 
> maybe i'm completely wrong here... are you world + kernel in sync? I would
> guess that you'd want them to be... maybe you can boot into 6 with a CD, mount
> all the disks as needed (assuming that userland-version-6 works fine with
> kernel-v6 and raid-version-5 ), finish installworld, and then reboot to start
> using the tools on the FS...

No, they're completely not in synch: I'm following the UPDATING
recipe, and am getting to the point after installkernel
where you have to boot to single-user mode to do the make
installworld.  That, obviously, has a _6 kernel (and
ataraid module) and _5 user-land.  The kernel has options
COMPAT_FREEBSD5, so that this step works: that's part of the
recipe.

Is there any particular reason that fsck or mount_ffs
from _5 would not work with _6?  That would seem to be a
prerequisite for the upgrade-in-place procedure.  I've done
an upgrade-in-place before, on my workstation, and that went
without a hitch.  The only difference (besides version creep
over the intervening six months or so) is that this is on a
server server with an ataraid RAID1 mirror array, rather than
raw disks.

Cheers,

-- 
Andrew