Strange Behavior Upgrading Dragonfish 24.04.2.5 to ElectricEel 24.10.2

TrueNAS General
1

Hi all,

We recently upgraded to Dragonfish 24.04.2.5, which went well. We then upgraded to ElectricEel 24.10.2 and have observed some strange behavior.

We tried to apply the ElectricEel upgrade once and it seems it didn’t take properly, so we applied it a second time. On the second attempt, the upgrade took and booted correctly. Everything seemed fine and then 24 hours later the console was in a strange state refreshing itself constantly, the shares were not accessible and the UI was hanging.

I rebooted into ElectricEel and SCALE was reporting a kernal taint with the CDDL (?) module and would not proceed beyond that.

I was able to reboot into Dragonfish, which also looked like it reported a kernal taint, perhaps with the same module, but that didn’t impede the boot process and it came back up.

I checked the boot environments in the UI and found the following.

There are two boot environments for ElectricEel with differing file sizes. It seems the first did not successfully complete?

While we understand what is happening, we’re going to stay on Dragonfish. We are going to mark the Dragonfish 24.04.2.5 as Active boot environment so that it boots up to Dragonfish on next reboot.

Any help interpreting what I’m seeing and why ElectricEel is acting strangely would be greatly appreciated.

Thanks.

Jorge

2

Image attached.

image
image2078×314 24.4 KB

Hopefully the explanation is helpful.

Thanks again.

Jorge

3

I suspect you are running into this Clean install of 24.10.0 results in "cannot import ‘boot-pool’ : no such pool available" - #14 by CubedRoot issue.

The kernel taint is normal as ZFS is licensed under CDDL and not GPL.

Does the console drop into an (initramfs) shell? See what I did to fix that here: TrueNAS SCALE 24.10 Failed to Find boot-pool After Installation – PK1048

4

Thanks PK1048. Sounds like it because we do have mirrored boot drives.

It’s odd because ElectricEel did boot initially and was stable and working correctly, and then this issue presented itself and it just halts at that kernel taint prompt and doesn’t drop to shell, unfortunately. If it doesn’t drop to shell, any ideas?

Great blog. Will read the entry and follow going forward.

I see the same kernel taint in Dragonfish boot, so your comment on CDDL makes sense. Appreciate the clarification.

We’ll be staying on Dragonfish for now on our production environment since this seems to be a larger issue. On our next maintenance window we may try to upgrade to ElectricEel again and remove the previous 24.10.0 instances from the boot environment.

5

If I remember correctly, the taint message happens just before it tries to import the boot-pool. Do you see any message on the console about importing boot-pool, if not, then it is hanging before that step in the boot process and I have no idea what it might be hanging on.