Kernel panic after update to 25.10.3

Hi, i tried to upgrade my truenas server to version 25.10.3. But after reboot i get a kernel panic. I am able to select the older version during grub, and it boots normally. I tried a manual upgrade, and this also failes. Somebody any ideas?

regards

Jack

We would need more details like your hardware setup. Only other suggestion is to boot to a previous boot environment and delete the 25.10.3 boot environment and try upgrading again.

You really are not giving us much info to work with.

1 Like

Hi, it is a bare metal box booting from ssd. Been running for a few months now. I was not able to remove the failing boot environment from the gui. I tried to reinstall but this adds a new boot environment which does also cause a kernel panic. I have another bare metal box which i use for backup. I tried the update here first without any problems. So i went for the upgrade on my main box. So was not expecting any problems

Did you go into the Advanced Settings, Boot section and make the working boot-environments active? I am guessing the 25.10.3 environment is set for reboot or something and that is why you can’t delete it.
Are there any messages that may give a clue as to where it is failing in the process?

Hi, the 25.10.3 environment state was restart. That was the reason it could not be deleted. But I fixed the issue. The boot pool was degraded. This happened during installation a while ago. There where no errors, but the state was still degraded. I cleared the state using zpool clean. After this the reboot still caused a kernel panic. I booted in the working version, tried the manual update again, and this fixed the issue. I was able to delete the version which was causing the issue. So for now the issue is fixed. Thanks for your reply! I am still kinda new with this, still a lot to learn :wink:

Using zpool clear just makes the system forget about the error, it does not resolve any errors.
If whatever caused the issues remain, the errors will soon return.

If your boot envirments are getting corrupted my guess as to cause would be faulty RAM (run memtest86), a failing boot device (S.M.A.R.T may give an idea of the current status) or a faulty Power Supply, in that order.

1 Like

Absolutely. If there are any hardware issues clearing the error does not solve anything. In this case it enabled the update. I will keep a close eye on it, and will run the suggested tests. Thanks for the tips!

Opinion: Run memtest86+, the FOSS version, in continuous loop for 5 days (!) to be reasonably certain consumer RAM is ok. Memory errors can be sporadic, and even a day or two of running is not nearly sufficient.

Some people have found errors after 3 weeks. I soaked my backup system for about 3 weeks. My main has ECC RAM because of this, I don’t want to be chasing Schrodinger’s bad RAM.

This brings back a request that’s been made a few times to the iX devs: include memtest86+ as a boot option on TrueNAS (as some other OS/distros do, Proxmox notably).

1 Like

Amazing QoL for home users on consumer RAM. As useful as wheels on a fish for Enterprise and prosumer setups bcs ECC. Let’s not start the ECC wars again and doooo eeeet.

Still no errors on my boot environment, everything is running without any issues so far. I am a home user using consumer RAM. ECC would be overkill at this moment (and current prices!) for the purpose of my system. The real issue was that my system would not boot a working environment without user intervention after the latest update. The zpool-clean command fixed this issue for me. Maybe not best practice, but for now this made my server usable again. ECC memory would be a no brainer if this system was a enterprise solution.

@breikers
I notice that you still have not provided a detailed description of your hardware setup as has been previously asked. It is difficult to offer much advice without an understanding of the problem.

Hi, not sure if this is helpfull beside the already provided information. But here is some more info:

AMD A10-7850K Radeon R7, 12 Compute Cores 4C+8G

16 GB Consumer RAM

Single 128 GB boot drive ssd.

500 GB mirror nvme for application datasets

1 x RAIDZ1 | 5 wide | 3.64 TiB Main storage.

Please let me know if you need any more information. At this time the problem has been resolved as said earlier.