John Childermass
Dabbler
- Joined
- Oct 20, 2016
- Messages
- 34
I posted this originally at the end of the thread mentioned below, but realised that as it was under the FreeNAS (Legacy Software Releases) section and I'm using TrueNAS Core I should probably post it here instead.
I currently have TrueNAS Core 12.0-U8 running with a single RAID-Z1 pool of 3 x 1TB drives. I want to change everything over to a single mirrored pool of two new 12 TB drives, after which I would retire the old 3 x 1TB drives or use them elsewhere.
As I set the original pool up just to get started learning about TrueNAS, I don’t have a lot of critical data on it yet. I have a single SMB share in one dataset (with no important data), and the virtual disks of a couple of VMs in another dataset. The VMs would be a little trouble to recreate, but they contain nothing critical. I’m therefore also wondering whether it would be simpler just to create a new pool with the new drives, recreate the SMB share on that pool, individually replicate the VM zvols to the new pool if I don’t want to recreate them, and move the system dataset over. I presume this would work too?
I currently have TrueNAS Core 12.0-U8 running with a single RAID-Z1 pool of 3 x 1TB drives. I want to change everything over to a single mirrored pool of two new 12 TB drives, after which I would retire the old 3 x 1TB drives or use them elsewhere.
- Are the instructions at the start of the thread "Howto: migrate data from one pool to a bigger pool" from @depasseg still the best way to do this?
- The existing pool is not encrypted (if I understand correctly, what I should really say is that none of the datasets in this pool are encrypted). If I want to use encryption after I move everything to the new drives, will @depasseg’s method work, or will zfs replication mean that the copied datasets on the new drives are unencrypted, even if I initially set up the destination pool as encrypted?
As I set the original pool up just to get started learning about TrueNAS, I don’t have a lot of critical data on it yet. I have a single SMB share in one dataset (with no important data), and the virtual disks of a couple of VMs in another dataset. The VMs would be a little trouble to recreate, but they contain nothing critical. I’m therefore also wondering whether it would be simpler just to create a new pool with the new drives, recreate the SMB share on that pool, individually replicate the VM zvols to the new pool if I don’t want to recreate them, and move the system dataset over. I presume this would work too?