Core to Scale, starting from scratch

TrueNAS General
1

I decided it’s time to migrate to Scale. I am on TrueNAS-13.0-U6.7 with 2 mirror pools of 2 drives each. One of the datasets is encrypted (not GELI as I remember). It is just a NAS, no jails or VMs.

I was thinking to start from scratch. Install Scale on a new disk, then import the pools and start making users, permissions, shares etc.

Could you verify me a few things:

  1. my datasets with their data and structure and their snapshots as well will be intact on the new machine?
  2. I have to export the pools on the “old NAS” first?
  3. I will make new snapshot tasks, and later I will be able to manually delete the snapshots from the previous configuration?

and last and more important

I will do the migration preparation on Core as described on the TrueNAS site and also I will do one last configuration export of the Core system.

In the few first days after using Scale if I find it too time consuming or have difficulties to make shares, permissions etc from scratch, should I be able to load the configuration file and continue from there?

Do I miss something critical?

2

yes.

A clean shutdown and import in the new system should be enough. For good measure you can export it.

yes

yes

2 Likes
3

Something that puzzled me a bit at the time of the sidegrade from Core to Scale (using the Core config file) was the match-mismatch of users/group.
Working by IDs, you can upload your config and

  • lose prebuilt users/group of the fresh install
  • Core users that have the same ID of prebuilt, instead will take theyr place

Off course nothing that cannot be fixed manually (and is not so huge work), i have to admit that at beginning i neither realize anything, just am hindsight share experience

1 Like
4

thanks for your answers!

What a noob I am! I forgot about datasets/files permissions and ownership. I’d like to make new users and start fresh.

Previous setup is done with ACLs, I searched a little and found that it is a bigger deal (for me) than I thought.

I’m between

stripping ACLs and starting users and permissions from scratch (I hope to not mess it)

and

formatting the pool and copying the files from backups (all of the data exist in backup disks and the most important datasets and their snapshots exist in a single USB disk ZFS pool)