We have toaster A, a F520 with 2 SCSI controllers and 3 disk shelves of 4 GB disks.
We may someday have toaster B, a F520 with one SCSI controller and no disks.
Potential plan: buy a shelf of 9GB disks, put it on toaster A, move a shelf of 4GB disks to toaster B, keeping all of the toaster A data on toaster A. Toaster B ends up with an empty file system on its 4GB disk shelf.
Now the mechanics of the move are a little unclear to me. Presumably we hook up the 9GB shelf to toaster A and 'raid fail' each 4GB disk in the shelf we want to move? But there will be 5 9GB data disks, since we have to burn one as a hot spare and one becomes the new parity disk. Will there be a problem with having 7 disks to move data from, but only 5 disks to move data to? There would be enough GBs on the 5 9GB disks, but...
The problem is if you have a 4GB data disk, and a 9GB hot-spare disk, if you fail the 4GB disk, it will replace it with the 9GB spare, but it will only make 4GB of the 9GB accessible.
Sounds stupid, but it's one of the few really bad weaknesses of the netapps. You can add new extra disks at the full 9GB capacity, but if they replace a "failed" disk, you only use the capacity of the original "failed" disk.
Don't forget that one of those 9GB disks will be the parity disk, no matter how you work it. If you want an online hot-spare, you'd better make one of the 9GB ones a hot-spare as well.
You always have the "full dump; reinstall os; full restore" option of course.
Perhaps you could use that second netapp for a new install: put the 5 9GB disks there. That give you 36GB. Perhaps you could then copy your data over. Then reinstall the first box with a smaller number of 4GB disks, and add the leftover 4GB disks to the second netapp (moving one of your scsi controllers if necessary). Swap hostnames. Reboot all clients (if necessary to clear stale NFS file handles). You're done.
Darrell Root rootd@nas.nasa.gov