So, how does this work logistically? Maybe the following:
Create a primary pool of 10TB + 3TB
Create a backup pool of 10TB + 3TB
Create a mother of all pools containing just the backup pool
Copy all information from primary pool to mother of all pools
Delete information from primary pool and add it to the mother of all pools
Enable duplication on mother of all pools
How does this help with backup solutions that don't like pools? Do you go down to the underlying disks themselves in the backup pool and just back them up?
Right now, I have a single pool consisting of all the HDD's with pool duplication turned on. If a 10 TB HDD dies, I should still have all of my files on the remaining drives but duplication stops if I have more than 8TB of information in the pool.
10TB Drive dies: 26TB - 10TB = 16TB/2 = 8TB max of duplication
3TB drive dies: 26TB - 3TB = 23TB/2 = 11.5TB max of duplication
Unless I'm missing something, it doesn't really matter with drivepool where the data resides. If duplication is turned on, I can lose a drive and still have all of my data. The bad drive just has to be removed from the pool and another of equal or larger size is needed to replace it (or multiple smaller drives adding up to the size of the drive lost). I guess I just need to stop being hung up on the RAID paradigm.
What's a good backup program that works with a Drivepool volume rather than having to go to the base disks?