25.10.2.1 - Goldeye Docker NVIDIA Driver issues

[gswhiteuk]

Under 25.10.2.1 - Goldeye, I am experiencing some issues with Nvidia drivers not loading and Docker containers failing to load due to the Nvidia errors. All was working prior to the upgrade to 25.10.2.1

I have run some commands within the shell

“lspci | grep ’ VGA ’ | cut -d" " -f 1 | xargs -i lspci -v -s {}
01:00.0 VGA compatible controller: NVIDIA Corporation GP107GL [Quadro P620] (rev a1) (prog-if 00 [VGA controller])
Subsystem: NVIDIA Corporation GP107GL [Quadro P620]
Flags: fast devsel, IRQ 77, IOMMU group 9
Memory at f5000000 (32-bit, non-prefetchable) [size=16M]
Memory at e0000000 (64-bit, prefetchable) [size=256M]
Memory at f0000000 (64-bit, prefetchable) [size=32M]
I/O ports at f000 [size=128]
Expansion ROM at 000c0000 [disabled] [size=128K]
Capabilities:
Kernel modules: nouveau, nvidia_drm, nvidia

➜ ~ nvidia-smi -L
NVIDIA-SMI has failed because it couldn’t communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.

root@elsa[/mnt/.ix-apps/app_configs]# nvidia-smi -q | grep UUID
root@elsa[/mnt/.ix-apps/app_configs]#”

➜ ~ midclt call app.gpu_choices | jq
{}
➜ ~

I have also tried following numerous threads on the forum, trying to uninstall Nvidia drivers and then re-installing. Unsetting the pool and then resetting.

Nothing appears to have worked. I use Emby for media streaming, which was utilising the Nvidia card for Hardware transcoding, but the Docker will not load due to the card/drivers no loading or being recognised in the system. Error message below:

“[2026/03/06 13:04:05] (ERROR) app_lifecycle.compose_action():58 - Failed ‘up’ action for ‘emby’ app: Network ix-emby_default Creating\n Network ix-emby_default Created\n Container emby Creating\n Container emby Error response from daemon: unknown or invalid runtime name: nvidia\nError response from daemon: unknown or invalid runtime name: nvidia\n”

Appreciate anyone’s time, advice, or help trying to help me fix this.

I have also read and tried this page in the forum, but as the drivers have not been loaded and my Nvidia card is not working, I am unable to get the UUID

Docker Apps and UUID issue with NVIDIA GPU after upgrade to 24.10 or 25.04

Morning all, We’re tracking an issue with Apps that impacts NVIDIA users, which we’ve now added to the Known Issues page of our release notes. Some users who have upgraded to 24.10 or 25.04 from a previous version, and who have applications with have NVIDIA GPU allocations, report the error Expected [uuid] to be set for GPU inslot [<some pci slot>] in [nvidia_gpu_selection]) (see NAS-132086). Users experiencing this error should follow the steps below for a one time fix that should not need t…

After reading many threads, I am at a loss on how to fix this, please help

thanks in advance

Gary

[LarsR]

Did you use a prior version of 25.10 or did you update from 25.04 to 25.10? And what is your gpu?
25.10 updates the nvidia driver and sadly any gpu older then 16XX cards is no longer supported by the newer driver.

Edit: just saw in your post its a quadro p620 which i believe is no longer supported by the newer driver. You could roll back to 25.04 and use the work around a community member created for older gpus and 25.10

NVIDIA compatible driver test for TrueNAS 25.10 “Goldeye”

TrueNAS 25.10 now uses the NVIDIA open GPU kernel modules with the 570.172.08 driver. This enables TrueNAS to make use of NVIDIA Blackwell GPUs - the RTX 50-series and RTX PRO Blackwell cards - which many users have requested support for. The NVIDIA 50-series Blackwell cards require the use of the new open GPU kernel module, but several of NVIDIA’s older generations of GPUs - including the Maxwell, Pascal, and Volta generations - lack the GPU System Processor (GSP) module on their silicon in or…

Edit2: For more information on the issue

NVIDIA Kernel Module Change in TrueNAS 25.10 - What This Means for You

With the upcoming release of the TrueNAS 25.10 “Goldeye” beta, there have been some changes made to the NVIDIA drivers and supporting modules that are compiled and shipped with TrueNAS. TrueNAS 25.10 now uses the NVIDIA open GPU kernel modules with the 570.172.08 driver. This enables TrueNAS to make use of NVIDIA Blackwell GPUs - the RTX 50-series and RTX PRO Blackwell cards - which many users have requested support for. [1] [2] [3] Unfortunately, what NVIDIA giveth, NVIDIA taketh away. The N…

[leoqtwyec]

The Quadro P620 is a Pascal generation card, and TrueNAS 25.10 Goldeye switched to NVIDIA’s open kernel modules with the 570 driver, which dropped support for older architectures like Maxwell, Pascal, and Volta. So the card itself is the problem, not anything he did wrong. LarsR’s advice to roll back to 25.04 is probably the most practical short-term fix, and there’s apparently a community workaround floating around for running older GPUs on 25.10 if he wants to stay on the newer version. Either way, long-term he’d need a newer GPU to stay current with TrueNAS updates.

[gswhiteuk]

Thanks to you both for your replies.

Shame my GPU is not supported worked so well

I am looking for the roll back software, can’t seem to find it on this site, the links for downloads appear to go in a loop?

Ig I ca continue support for the card inside 25.10 that would great, is there a forum link I could follow please?

much appreciate your help and advice.

[SmallBarky]

You roll back your version of TrueNAS by selecting a previous boot environment. System, then Boot and select a previous version. Once booted in the old version, you can go back to that menu and delete the newer boot environment.

[LarsR]

@gswhiteuk

Ig I ca continue support for the card inside 25.10 that would great, is there a forum link I could follow please?

the first link in my first reply is the post with the workaround

[gswhiteuk]

Perfect thanks I will take a look.

Thanks again