Is it true that all Fangtooth instances must be in the same pool?

Apps and Virtualization
1

I have Scale Electric Eel 24.10 installed on a custom server with a 12TB HDD pool and a 4TB SSD pool. I have 6 VMs, three of which are on the HDD pool and three of which are on the SSD pool.

I am planning for the upgrade to Fangtooth, and when reading the instructions for manual migration of Electric Eel VMs to Fangtooth instances, it looked to me as though all instances must be in the same pool in Fangtooth.

Is this correct? In particular, is it true that all Fangtooth instances must be in the same pool?

If so, then I’m going to need to buy another SSD and expand my SSD pool before upgrading to Fangtooth, because some instances need to be on SSD, and I’m not going to be able to fit all instances in a 4TB pool.

2

Yes, there’s a single .ix-virt dataset for Incus.
But why are your VMs so large? Are you storing data in the VMs which should rather be in datasets on the HDD pool and then mounted into the VM?

1 Like
3

Thanks for the clarification. That’s too bad, but it’s what I suspected.

One of the VMs is a Nextcloud server, which has a handful of users and 1 TB capacity to ensure that users have space for their data, even though it’s not that full currently.

Another VM is a Linux Mint machine I use for manipulating and compressing large files and folder structures. It’s also 1 TB capacity, even though when I’m not actively using it for data crunching, there is very little in it.

There is also a Windows machine and a few other VMs. All told, around 3 TB of capacity in the six VMs, So technically I could just fit them in a 4 TB pool, but it wouldn’t leave much headroom or room for snapshots.

4

Look into moving user data out of the VMs.

1 Like
5

The only user data in a VM is in the Nextcloud server. That’s an interesting thought, but the Nextcloud setup is more complicated when external storage is involved. Still, it’s not a bad thought, and I will consider it. I will also add that, currently, the Nextcloud data is mirrored every half hour to a Truenas dataset using rsync, and from there to Backblaze. So, the data is protected.