*scratches chin*
OK, so you have to take some care if your volume group extends across multiple physical drives. In my case, I run pairs of drives in RAID 1 and use them to feed my volume groups.
Code:
~# cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] [multipath]
md1 : active raid1 sde3[1] sdd3[0]
142716800 blocks super 1.2 [2/2] [UU]
md2 : active raid1 sda3[2] sdc3[0]
974999360 blocks super 1.2 [2/2] [UU]
md0 : active raid1 sde2[1] sdd2[0] sdc2[2](S)
523968 blocks super 1.2 [2/2] [UU]
unused devices: <none>
~# pvs
PV VG Fmt Attr PSize PFree
/dev/md1 mdgroup lvm2 a-- 136.09g 224.00m
/dev/md2 mdgroup lvm2 a-- 929.81g 84.91g
In my case, I've got 4 hard drives that look like 2 hard drives that provide storage for my one volume group. I can lose one drive from each RAID 1 array and not lose anything (assuming that I get off my butt, buy replacement drives, and move the physical extents on the degraded RAID array to another physical volume before the other drive in the array craps out). (/dev/md0 is my /boot partition, in case you're curious.)
That isn't the only way to do it, but it was simple enough for me to figure out and use.
Now, if both drives in /dev/md1 were to fail
at the same time, I'd lose ~136 GB of information. That physical volume would be gone and any physical extents on it would be lost to me (unless I had a decent backup somewhere). I wouldn't even know what exactly I had lost, since the physical extents that I had lost could be in any logical volume (depending upon how and in what order I had created/grew them).
In *your* case, you should keep each drive in its own volume group just so that you can control where the real storage for your logical volumes are going to end up. If you ever decide to upgrade your drives, you can add the new drive to the volume group of the drive you intend to replace and then issue a pvmove command to have the physical extents move off the old drive to the new one while the system is running. Once pvmove tells you that it is done, you can remove the old drive from the volume group and then remove it from the system when convenient. You can use the unallocated extents in a given volume group to grow or create new logical volumes as needed.