Pools Break When Upgrading from Core to Scale (Dragonfish)

TrueNAS General
,
1

I’ve used TrueNas for some years now. Upgraded to TrueNas Core from FreeNas way back when and now hoping to migrate to Scale. The problem I’m having is when I attempted to migrated from Core to Scale, upon boot, Scale could not identify two out of my three pools.

I looked at the disks the system could see and tried zpool import. It showed me that my data pool could see two of the five disks with the remaining disks showing as unavailable and thus the pool could not boot.

In some of my readings, some folks had indicated that there could be old encryption from a long long ago install that could be breaking the pool. I’ve looked and I believe all the drives and pools are unecrypted.

All this to say, I’m at a bit of a loss of what to do next. Everything I’ve seen thus far made it seem like this process was simple and straightforward. Especially for just simply migrating the pools and VDevs.

What is my next step here?

After all this I got concerned and booted back to Core from the boot options that pop up and my pools were recognized no problem by Core. however, I’m now getting a disk error along the lines of:

“ValueError: 30 is not a valid PoolStatus”

If anyone could offer me any guidance I would greatly appreciate it. At this point, I’d be happy to forego Scale until things get ironed out and just stay with core, though, now with this new error, I can’t view my disks in the GUI.

Thank you in advance for any advice and let me know if there is any additional information I can provide to help troubleshoot.

2

While in Core, what doe this command reveal?

geli status

You’re better off issuing that command in an SSH terminal, rather than using the broken “Shell” in the GUI.

3

Screenshot 2024-05-22 215501

Hi Winnie, thank you for the reply, this is what I get returned with that command.

I’m not sure exactly what it means though.

4

What is your hardware? Specifically what are your disk controllers?

What version of SCALE did you try to upgrade to?

1 Like
5

I’m using an LSI 9218 SAS controller attached to 8 disks and an older Phenom x2 965 Black Edition processor from way back when. There’s another 6 or 8 disks connected to the motherboard.

I attempted to upgrade to Dragonfish.

One of my observations is that some folks who were even on Cobia were having issues with getting their pools recognized with their upgrade to Dragonfish. Is it possible that there is an issue with Dragonfish itself? Would I be better served upgrading to Cobia instead and trying that?

Currently I’ve used the GNU boot selection to start back from Core, but it looks like some things broke specifically because I can no longer view my disks from the GUI. It gives me the error: ValueError: 30 is not a valid PoolStatus

My next step was I bought a large drive that I’ll back everything up to as well as get a new boot drive. I’d pull my current mirrored boot drive out and install Dragonfish on that and then upload my configuration and see if my pools show up.

6

So a decent HBA. There is an issue with some missing drivers but not for this.

If you have watched the forums there are loads of people saying that they have issues with the first release of Dragonfish.

All new versions have bugs, but my impression is that Dragonfish has more than Cobia.

Cobia is stable - there are some issues with GUI reporting that won’t be fixed until Dragonfish but not AFAIK with pools.

So personally my general advice (as a non-expert) is for people to avoid Dragonfish for existing systems unless they don’t mind a possible abortive upgrade, and to stick with Cobia for the moment. Many of the reported pool problems do not seem to have fixes in 24.04.1 which is imminent, so personally I think we need to remain patient and see what ixSystems does about all the people reporting pool problems.

ixSystems position on upgrades can be found on their Software Status webpage and effectively I am saying that existing production boxes should follow the Conservative recommendations on this table.

See also the question I asked in When to upgrade to Dragonfish? and feel free to add your own issues there.

1 Like
7

Given that I’ve already kind of started the upgrade process, should I try to re-upgrade to Cobia and try that? Or is my system kind of borked as it is?

I’m getting a new disk this evening that I’ll start on a task to back up everything from the server, but after that, I suppose I’m flexible.

Would it work for me to attempt another upgrade to Cobia instead of Dragonfish? Or would I just bork my system more than it already is?

8

It might work. What won’t work is downgrading dragonfish to Cobia.

if you backed up your core config (and you should’ve been prompted to) then you should be able to clean install Cobia and upload your core config.

9

So, technically, when I start up my boot option is set to CORE. But there’s an option for Dragonfish. I’m currently in Core now, it looks like I can select Cobia as an update/upgrade path. I am assuming since I started this process with Dragonfish already it’ll bork my install? I figure it’ll just be better to reinstall with a new system and go from there?

Do from scratch Dragonfish have pool issues or just from updated/upgraded system paths?

10

I think it’ll work actually.

11

Sweet, once I backup all my stuff I’ll try it and see what happens. Then I’ll report back with my findings. Thank you all for your help and advice.