I previously had a 500 GB cache drive (single disk pool) and replaced it with a 1 TB drive.
I shut down docker and virtual machines, set shares to Cache: Yes and invoked mover. After mover finished I took the 500 GB drive out and put the 1 TB drive in. I flipped the shares to Cache: Prefer and invoked mover. Mover has been going for over 24h now and it has somehow managed to move over 500 GB of data back to the 1 TB drive - which makes no sense since the 500 GB drive was only about 45% full to begin with.
Any idea why this is occurring? It appears to be related to btrfs subvolumes - it has been moving them for almost the entire time and somehow growing them to bigger than they were previously.