> Any ideas on how I can remove the LV and PV for the second disk from the
> VG so I can get my life back in order?

Looking at my fedora core 3 init script, LVM2 setup boils down to this:

modprobe dm-mod
echo "mkdmnod" | /sbin/nash
vgscan --mknodes --ignorelockingfailure
vgchange -a y --ignorelockingfailure

Either its vgscan or vgchange thats failing. My bet is on vgchange. That
should be okay. If you're getting past the vgscan part, you should be
able to use the lvm tools to fix things. vgreduce even has a --
removemissing command that "Removes all missing physical volumes from
the volume group and makes the volume group consistent again."

If vgscan is barfing, well, see if there's an option to ignore missing
disks.

Back when I was using LVM1, I ran into a similar SNAFU. My root wasn't
on LVM. I ended up copying the last metadata backup in /etc into where
ever it was the current metadata goes, which is normally set up by
vgscan. The lvm tools then worked. But if your root is on LVM, you've
got a catch-22 on your hands...

... I was just about tempted to put my root partition on LVM, but you
just reminded me why its probably a bad idea.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://shadowknight.real-time.com/pipermail/tclug-list/attachments/20050403/ecec5871/attachment.pgp