GPU is not showing on the Plex hardware acceleration setting

[RealMoto]

Hello,
I am not sure if this has been covered somewhere else, but I could not find it. I set up plex on my Truenas scale and it's running. But I was not able to add my GPU to the resource reservation setting. My GPU is not coming up as a choice.

I was able to go under the system settings & advanced settings to the Isolated GPU device and select my GPU.

Is there something I am missing? I am new to all of this and willing to learn.
For now, all I am trying to use is a Plex server.

Thank you for your assistance,

 

[RealMoto]

I am not sure if this helps, but I took a screenshot of this.

 

[whodat]

Hi @RealMoto , assuming you are trying to set up Plex as an 'App', the GPU Isolation feature is for the sake of VM's only, not Apps.

Try removing the Isolation, then reboot and see if your GPU is available in your Plex App's settings for 'Resource Reservation'

 

[RealMoto]

Hi @RealMoto , assuming you are trying to set up Plex as an 'App', the GPU Isolation feature is for the sake of VM's only, not Apps.

Try removing the Isolation, then reboot and see if your GPU is available in your Plex App's settings for 'Resource Reservation'

Thank for the help.
I did try your suggestion but no luck. On the Plex app setting for Resource Reservation, the GPU dose not come up.

I did what was suggested and removed the GPU from the Isolated GPU.

Is there something else I can try?

 

[whodat]

I am not sure if this helps, but I took a screenshot of this.
View attachment 59108

Where did you run these commands? Which shell/ CLI are you in?

Try running these commands from WebGUI > System Settings > Shell , and comment back with the output please.
(I have included my own output for reference)

Code:

# nvidia-smi
Fri Oct 14 21:32:12 2022
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.103.01   Driver Version: 470.103.01   CUDA Version: 11.4     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  Off  | 00000000:01:00.0 Off |                  N/A |
|  0%   38C    P8     8W / 130W |      0MiB /  5944MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+


+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+

Code:

# lspci -v |grep NVIDIA
01:00.0 VGA compatible controller: NVIDIA Corporation TU116 [GeForce GTX 1660] (rev a1) (prog-if 00 [VGA controller])
01:00.1 Audio device: NVIDIA Corporation TU116 High Definition Audio Controller (rev a1)
01:00.2 USB controller: NVIDIA Corporation TU116 USB 3.1 Host Controller (rev a1) (prog-if 30 [XHCI])
01:00.3 Serial bus controller [0c80]: NVIDIA Corporation TU116 USB Type-C UCSI Controller (rev a1)

 

[RealMoto]

Where did you run these commands? Which shell/ CLI are you in?

Try running these commands from WebGUI > System Settings > Shell , and comment back with the output please.
(I have included my own output for reference)

I ran the command from the web GUI (System Settings) shell.


I tried both codes you provided to me, but I didn't get the same results.

I can be doing something, I am doing wrong.

 

[whodat]

Run the nvidia-smi and lspci -v | grep NVIDIA commands from the root@truenas[~]# prompt.

Don't enter cli before running them.

 

[RealMoto]

I was able to get this info.

 

[whodat]

I suspect that your Quadro M2000 is not supported in TrueNAS SCALE. Unfortunately, I am not aware of a list which shows the GPU's that are supported.
Looks like you already found this thread, which seems to confirm the same:

Nvidia GPU not appearing for use with SCALE

I installed a Quadro 2000 (2014 model) GPU into my main server today running TrueNAS-SCALE-22.02.1 and expected it to show up in the GPU section of Apps so that I can use it with iSpy (performance is terrible in iSpy and I am hoping this will help). But, no matter what app container I try to use...

www.truenas.com

 

[RealMoto]

Run the nvidia-smi and lspci -v | grep NVIDIA commands from the root@truenas[~]# prompt.

Don't enter cli before running them.

I ran the commands as you mention and these are the results. It's contradicting one command that says No devices were found, but the other gives a result with the device.

 

[AndyMcSchopf]

I suspect that your Quadro M2000 is not supported in TrueNAS SCALE. Unfortunately, I am not aware of a list which shows the GPU's that are supported.
Looks like you already found this thread, which seems to confirm the same:

Nvidia GPU not appearing for use with SCALE

I installed a Quadro 2000 (2014 model) GPU into my main server today running TrueNAS-SCALE-22.02.1 and expected it to show up in the GPU section of Apps so that I can use it with iSpy (performance is terrible in iSpy and I am hoping this will help). But, no matter what app container I try to use...

www.truenas.com

This i can deny - I am using one M2000 in my Server (and a P2000 in the other) and it is (both are) working flawless. Using it in an Plex-Container.

 

[whodat]

Based on @AndyMcSchopf's screenshot, by comparison it looks like your NVIDIA M2000 is "active" @RealMoto i.e. I think it is being used as the host's console and is therefore busy and not available to Apps?

See the note on "Any controller with [VGA controller] at the end " here:

How to check which GPU is active in Linux?

I have 2 GPU's in my netbook. How do I know which one I'm actually using at any given moment?

unix.stackexchange.com

Are you able to set the primary graphics device to a different VGA controller in your BIOS? Or otherwise try booting without a screen plugged into any of the VGA / DVI / HDMI/ DP ports on the M2000?

 

[RealMoto]

This i can deny - I am using one M2000 in my Server (and a P2000 in the other) and it is (both are) working flawless. Using it in an Plex-Container.


View attachment 59210

Thanks for the info. I was worried that the card was not compatible.

 

[RealMoto]

Based on @AndyMcSchopf's screenshot, by comparison it looks like your NVIDIA M2000 is "active" @RealMoto i.e. I think it is being used as the host's console and is therefore busy and not available to Apps?

See the note on "Any controller with [VGA controller] at the end " here:

How to check which GPU is active in Linux?

I have 2 GPU's in my netbook. How do I know which one I'm actually using at any given moment?

unix.stackexchange.com

Are you able to set the primary graphics device to a different VGA controller in your BIOS? Or otherwise try booting without a screen plugged into any of the VGA / DVI / HDMI/ DP ports on the M2000?

Okay, here is my setup. I am using a Dell T310 server. The monitor is plugged into the VGA on the server MB, and the GPU is in the PCI lane, but I don't have anything plugged into the GPU out ports. Here are some images. By the way, the reason I have the monitor is once the server boots I have to press F1 to lunch the Truenas server.

 

[RealMoto]

Based on @AndyMcSchopf's screenshot, by comparison it looks like your NVIDIA M2000 is "active" @RealMoto i.e. I think it is being used as the host's console and is therefore busy and not available to Apps?

See the note on "Any controller with [VGA controller] at the end " here:

How to check which GPU is active in Linux?

I have 2 GPU's in my netbook. How do I know which one I'm actually using at any given moment?

unix.stackexchange.com

Are you able to set the primary graphics device to a different VGA controller in your BIOS? Or otherwise try booting without a screen plugged into any of the VGA / DVI / HDMI/ DP ports on the M2000?

I followed the Link and the steps this is the message I received.

So, from reading the post looks like the GPU is in use, and that is why I can use it as an app. Looks like am getting some answers and making progress. So now how do I tell the server not to use the GPU?

I did what you suggested and unplugged the monitor. I still get [VGA Controller] for the Nvidia GPU.

 

[RealMoto]

So I added a second GPU to see what happens and ran the commands.

 

[whodat]

I did what you suggested and unplugged the monitor. I still get [VGA Controller] for the Nvidia GPU.

Are you sure? I don't see [VGA Controller] in the output of your lspci -v | grep -i nvid command in your screenshot here, after you unplugged your monitor.

That said, your nvidia-smi command still shows the GPU is not found...

By the way, the reason I have the monitor is once the server boots I have to press F1 to lunch the Truenas server.

Why do you need to press F1? Can this be disabled in BIOS, so the server boots TrueNAS SCALE without intervention and you can leave your monitor unplugged from boot?

Having seen the pics of your setup, it looks like your VGA is an onboard connector, rather than coming from your NVIDIA card. If you remove all cards, does the onboard VGA still work?

Also, do you have options in your BIOS to choose the active VGA controller?

 

[RealMoto]

Having seen the pics of your setup, it looks like your VGA is an onboard connector, rather than coming from your NVIDIA card. If you remove all cards, does the onboard VGA still work?

Also, do you have options in your BIOS to choose the active VGA controller?

I unplugged the monitor from the onboard connector, but the same issue still exists.

For the BIOS, I have looked for a way to disable the GPU on PCI slot 2, but It doesn't have that option.