Is it worth doing the update the yet ?
been using Dragonfish for a few months don’t have any issues .
so was wondering is it safe to do the update to EE
only have Plex installed but really don’t need the trouble of configuring if it goes wrong .
I would take what iX say with a BIG pinch of salt. There are quite a few examples on these forums of people who have upgraded and then found some of their apps not working. If you are not technical enough to fix these, then you may regret the upgrade.
I am not upgrading my Dragonfish box until at least 24.10.1, likely 24.10.2 or possibly even 24.10.3, and I am a more technical TrueNAS user but very cautious about breaking things.
When you do upgrade make sure that you read the Release Notes and do the pre-requisite actions documented in there.
2 Likes
I’m similarly planning on waiting for a while. The question only you can answer is whether its new features are worth the possible disruption. The big features in EE are:
- Docker compose-based apps rather than Kubernetes in Dragonfish and earlier
- Lots of people are so thrilled by this that they’re rushing (or have rushed), lemming-like, into an upgrade even to a pre-release version of EE
- Some of them are seeing problems with their apps
- If you have any TrueCharts apps, they will not automatically migrate, and you’ll lose them if you upgrade to EE
- RAIDZ expansion
- Revamped GUI
At some point, you’ll want to upgrade. Is that point now? IMO, unless those features are really important to you, waiting is a better idea.
You can look at the software status to see whom iX recommends the different versions to.
Do you see yourself as a “Tester”?
My recommendation is that you wait a bit.
Each to their own… but from the older controversy of applications in Core, to jails breaking on upgrades, to buggy virtualization in bhyve, to Coral, to then iX introducing apps in Scale but seemingly continuing to break them with every major upgrade… I’ve been watching from the sidelines and in the meanwhile am very happy with TrueNAS as a NAS - but I can’t understand why anyone would put themselves through the pain of trying to use it for anything else. Maybe it will change with the product stabilising and maturing but track record so far isn’t good.
As mentioned above - if you’re using apps, hold off upgrading for a while. For vanilla NAS, 24.10 works fine and seems stable enough.
1 Like
I have to agree. The only non-vanilla thing that I haven’t tried yet is Active Directory integration, the two that I use this with are still running Core (current). But I do need to test AD integration with Scale at some point.
You can, but even that is often overly optimistic.
An alternative opinion,
Test the upgrade. If you have an issue report a bug and rollback.
The upgrade is designed to be able to be rolled back, even if you are using apps. Just reselect your previous boot environment in System Settings → Boot
If you don’t report the bug then it may not get fixed in the .1 update.
1 Like
Probably a good percentage of people using apps are using Scale as an all-in-one storage and media server for home. I don’t want to run a 2nd server for apps.
I upgraded last week, only had 1 app that had an issue and that was my fault (tailscale needed an updated auth token). Already using my own compose apps, iX & community seem allergic to VPN, so I did it myself. Now thinking about migrating all my iX official/community apps to my own docker-compose.
iX has also kept the datasets for k3s system & apps, so you can roll back to dragonfish if needed.
Great things so far:
- Don’t have to rely on other people to create Helm Charts, learn Helm & Kubernetes & roll out your own catalogue or try to get the custom app UI to do what a project needs. Or run/manage a VM or sandbox.
- Lower overhead with docker vs k3s. Going from a load of 0.3 to practically idle. Less heat, less energy and a more responsive system.
1 Like
Good that it works for you.
I can only echo this. Took the plunge the other day. The only annoying thing was ditching the IP addresses that I had given certain custom apps (because this isn’t supported in EE) and reconfiguring things to use port forwarding instead but I guess this simplified things from an IP management POV on my network (for people who have use-cases where different IPs must be used, I believe this is still possible, but you’ll need to learn about docker networking and do this manually).
Had to re-run the migration process for Paperless because this is using an encrypted dataset but after that all is good. As mentioned above, the fact that the system is under less load means energy usage and saving money in the long run.
I am holding off until 25.04. I am using jailmaker on 24.4 and need none of the new features.
25.04 will hopefully bring “native containers” so I can ditch jailmaker.
And I want separate IP per container. I use that for a few ones.
If you need none of the new features then I’d say don’t update.
I am submitting the above comment as an entry for the “Under-statement of the Year” competition.
Fair enough. But note that jailmaker sandboxes should still work, so you may consider upgrading to .1 or .2 later on.
You are absolutely correct but I’d rather wait for incus and migrate everything.
It’s “Tester” right now, which use type do you think is appropriate today with regards to EE?
Right now, I think Tester is probably appropriate. But Kris has said they intend to change that to “general use” within a week or so, while I wouldn’t recommend that until .1 at the earliest.
Thank you.
My takeaway from Kris’ post was that they would evaluate the status after a week.
Fair enough. It’d be more accurate to say I understood him to intend to make that change that soon; I think you’re right that he didn’t explicitly say that.
We are literally in the meeting to review the status page right now 
3 Likes