Migrating Core to Scale - any last advice

Hello everyone.

Longtime Freenas then Truenas Core user.

I upgraded a different machine, and now have an extra board and CPU I want to put into my TruNAS.

Right now I am running a 4-drive Z2 with TrueNAS core (13) on 2 mirrored 16GB thumb drives. It has been running for at least 6 years now with no issues, thorough updates and upgrades both :slight_smile:

Anyway, I upgraded a different machine, and now have an extra board and CPU lying around I want to put into my TrueNAS. I’m gonna take this opportunity to beef up the system a bit. I use it simply for home data sync, and backup, as well as media service. Currently I have plex, syncthing, a couple other lightweight iocages.

I wanted to bounce this setup off of you all and see if there are any pitfalls or advice you could share!

My plans-
1- move over to TrueNAS scale. I want Docker, I am tired of jails, and it seems like Core is old news. (is this true?)
2- Install it on mirrored 128GB M.2 SSDs rather than the USB thumb drives.
3- expand my pool to 6-drive Z2 by adding a HBA card (IBM H1110 9211 HBA )

The new system will be an i7 770 with 64GB of non-ECC (i know, I understand)

I was considering trying to migrate the pools, but right now am planning on backing up all the data on a portable drive, as well as its home locations on desktop machines and doing a fresh SCALE install. Replace all the iocage installs with new docker installs. Copy everything back in. And hopefully do little to no maintenance for another decade?

Now, I am a hardware person, please dont get too nerdy.
Should I be doing this differently?
I could squeeze another SSD in there for a cache drive - but RAM cache should be fine, right? Anything else I should do that is high-yield?

Thanks in advance!

If you are planning to migrate from TrueNAS CORE to SCALE, we strongly recommend reading and using this article as your guideline before you start the migration process.

We also recommend you follow the migration path documented here:

There are changes that occur between major releases you need to address before the final upgrade to the most current SCALE release.

1 Like

is migrating best? Or should I build fresh and copy data back over?
I intend to replace my boot pool with a different pair of disks anyway, and my user setup is not particularly complicated.

If I could keep my pools in place that would be cute, but my data is not uber critical, it’s all backed up on desktop machines, and I am currently making another backup on a spare drive. PLUS I intend to add disks to my pool, and I am reading that expanding a pool is a headache of its own.

You can migrate by either upgrading through the UI, observing the major release to major release path, or by clean installing from iso files. The migration preparation article is a guide to help you successfully migrate from CORE to SCALE and not miss something important.

ok ok ‘rtfm’ :slight_smile:
I did and I will!

Thank you for your time.

Not mentioned in the how-to-migrate article is SMB auxiliary parameters.

It is likely best to nuke yours altogether if you have ever set any (FreeNAS Time Machine, to eliminate DOS compatibility, etc). Otherwise, you may be faced with a SCALE GUI insisting that SMB is running but no ability to connect to SMB shares either. Only once you stop + restart SMB will the service admit that it cannot start.

To fix this issue in SCALE, you have to drop into the shell and do it from the CLI since iXsystems removed the auxiliary parameters menu item from the SMB settings GUI. It’s easier to nuke these settings in the advanced SMB CORE GUI before you migrate.

Kind of hilarious that CORE worked for years with invalid SMB auxiliary parameters. SCALE perhaps has a more delicate digestive system…

Heh. I got bitten by a similar issue, but with NFS when I migrated.

1 Like

I don’t mean to hijack this thread, but I think I understand that if you upgrade/migrate from Core to Scale, it is mainly for your pools/data. The biggest thing holding me back is migrating my jails to docker. Is there guidance to help make the virtual sandbox migration a bit easier? I currently put the config directories in a separate dataset on my pool, so I would assume I could just setup another instance of something like plex and then point it at the config directory right?

If you want to minimize downtime, you can setup an Ubuntu or Debian vm, install docker on that.

Migrate while on core.

Then migrate to scale.

Then migrate from the Vm to a sandbox on scale.

Easier than it sounds :slight_smile: