8 hours ago, ich777 said:

Do keep in mind that you will need UEFI for newer cards alongside with other requirements like resizable bar, which is also recommended for 40 series Nvidia cards, Above 4G Decoding and so on because otherwise they will most certainly work not like they should and could lead to other funky behavior.

 

Sorry if my answer came over offensive this was not my intention but I think you are running a pretty old Dell Server correct?

 

Just because I saw it, why are you running Palworld in a LXC container, I really can't imagine that this is any faster since Proton is only a fancy name for WINE with scripts/patches for some games and they usually not working that well for dedicated servers since Proton is optimized for the Clients so to speak the Games that you run and play on a local computer.

The dedicated Docker container is now stable and also the RAM usage is under control.

 

No worries, take your time.

So....... 

 

Figured it out, it was my user script for the power saving.

 

Isn't there something I can put in the "go" file to do that?

 

This is the script that was running.

 

On mobile right now so formatting is bad 

 

#!/bin/bash
#nvidia-smi --persistence-mode=1
nvidia-persistenced
fuser -v /dev/nvidia*

 

Isn't the "nvidia-persistenced" what we put in the go file?

 

Edit, forgot to say, looking to move the server to an AMD EPYC or newer Xeon here sometime and UEFI and Resizable Bar is a requirement for that next server.

Edited March 19, 20251 yr by AnnabellaRenee87

48 minutes ago, AnnabellaRenee87 said:

Isn't there something I can put in the "go" file to do that?

For the Nvidia Driver, just add this after the emhttp line:

nvidia-persistenced

 

This will do fine.

Please don't do anything with fuser! Just put nvidia-persistenced in the go file and that should do the job, it could take a bit after the reboot so that it drops to a lower power state (usually 1 to 2 minutes) but that works just fine.

 

Glad that you've figured it out!

 

50 minutes ago, AnnabellaRenee87 said:

Edit, forgot to say, looking to move the server to an AMD EPYC or newer Xeon here sometime and UEFI and Resizable Bar is a requirement for that next server.

I'm still on my i9-10900 and couldn't be happier about it since it has enough power for my needs, even when compiling the driver packages for the 3rd party drivers.

Hi, sorry to be a pain. I wondered if it was possible to have the latest 570.133.07 Open Source Kernel Module drivers on Unraid 7.0.1? Thanks

10 minutes ago, LiamP02 said:

Hi, sorry to be a pain. I wondered if it was possible to have the latest 570.133.07 Open Source Kernel Module drivers on Unraid 7.0.1? Thanks

Sorry currently I only compile the latest driver version from the open source driver which is available at the release from the Unraid version you are running.

This might change in the future but for now it is what it is, however if you are willing to upgrade to 7.1.0-beta.1 (which is stable since I run that on my production Server where I also compile and create the plugin packages for the 3rd party driver plugin) then you would be able to download this driver.

 

Background to that is that the plugin is in it's current state not fully able to detect the correct version if there are multiple version from the OpenSource driver in the repository available.

 

I hope that somehow helps, sorry for not having better news currently.

2 hours ago, ich777 said:

Sorry currently I only compile the latest driver version from the open source driver which is available at the release from the Unraid version you are running.

This might change in the future but for now it is what it is, however if you are willing to upgrade to 7.1.0-beta.1 (which is stable since I run that on my production Server where I also compile and create the plugin packages for the 3rd party driver plugin) then you would be able to download this driver.

 

Background to that is that the plugin is in it's current state not fully able to detect the correct version if there are multiple version from the OpenSource driver in the repository available.

 

I hope that somehow helps, sorry for not having better news currently.

Ah, that’s okay thank you. I’m having issues with a VM crashing in 7.1 that didn’t in 7.0.1, but I’ll look for another solution.

 

17 minutes ago, LiamP02 said:

I’m having issues with a VM crashing in 7.1 that didn’t in 7.0.1, but I’ll look for another solution.

Can you please explain what you mean with that?

If you are using the card in a VM please don't install this plugin because that can crash you whole Server, as pointed out on in the first post!

 

This plugin is only meant for users who are planing to utilize their Nvidia GPU in Docker containers or LXC containers.

I am looking for help here. I have been trying to get Plex to transcode for a couple weeks now. I have Unraid 7.0.1 with the linuxserver Plex docker. I do have plexpass. I installed this plugin and have tried every driver available on the list. My card is GTX1060 (6 GB) and am currently on driver 550.

 

I am getting FFMPEG CUDA errors (see below) indicating that "system not yet initialized" and "Codecs: hardware transcoding: opening hw device failed".

 

If anyone can come up with anything else I could try, I'm kind of at my wits end with this. Thanks in advance!

 

Plex config:

 

Here is nvidia-smi:

 

And Plex settings:

 

Whenever I try to use Plex to Transcode, I just get "transcode" (no "hw"), the nvidia-smi shows zero activity, and I get a whole bunch of "FFMPEG" errors indicating that Cuda isn't started or is otherwise unavailable when I view plex logs.

 

Plex Dashboard:

 

Plex error log:

ERROR - [Req#aa/Transcode] [FFMPEG] - cu->cuInit(0) failed
ERROR - [Req#aa/Transcode] [FFMPEG] -  -> CUDA_ERROR_SYSTEM_NOT_READY: system not yet initialized
ERROR - [Req#aa/Transcode] [FFMPEG] - 
DEBUG - [Req#aa/Transcode] Codecs: hardware transcoding: opening hw device failed - probably not supported by this system, error: Generic error in an external library
DEBUG - [Req#aa/Transcode] Could not create hardware context for h264_nvenc
DEBUG - [Req#aa/Transcode] Codecs: testing h264 (decoder) with hwdevice nvdec

nas-1-diagnostics-20250323-0045.zip

Edited March 23, 20251 yr by strider2112
added diagnostics, removed personal data

6 hours ago, strider2112 said:

I am looking for help here. I have been trying to get Plex to transcode for a couple weeks now. I have Unraid 7.0.1 with the linuxserver Plex docker. I do have plexpass. I installed this plugin and have tried every driver available on the list. My card is GTX1060 (6 GB) and am currently on driver 550.

Is this the first time you are trying to utilise your Nvidia GPU in a Docker container or was it working previously?

 

Please never runs such a container as Privileged, this is not a switch that will solve all your issues, exactly to other side is happening in your case because you are running the container in Host networking mode you basically give the container full access to your Hosts network interfaces and Privileged is always a thing that should be avoided at all cost.

 

6 hours ago, strider2112 said:

I am getting FFMPEG CUDA errors (see below) indicating that "system not yet initialized" and "Codecs: hardware transcoding: opening hw device failed".

Seems like it can't utilise your Nvidia GPU, please upgrade to the latest Nvidia Driver in the plugin, reboot and try to official Plex container and see if it's working with that one.

Maybe you have to install a docker mod into the Linux Server container but I'm not sure about that because I'm not familiar with their container.

 

Also make sure to enable Above 4G Decoding or in some cases it's named Support for Huge Address Spaces in the BIOS at the PCI submenu.

 

Please also double check if you don't have any additional spaces at the end of the UUID in the Docker template.

Everything seems fine from what I see in your Diagnostics.

 

ATTENTION: Please try forcing a transcode with a native Plex app like for iOS or Android since it a known issue that the Plex Web Player or Windows App just shows a black screen and transcoding is not working there in combination with Nvidia.

Thank you for your support and efforts @ich777! I'm going to try the things you suggested below. I might uninstall the plex, remove all the data, update the driver, then restart, then install the official app and see if that order of things helps anything. I've left comments on your comments and suggestions below. I'll report back once everything is up and running. Thanks again! Your dedication to this is inspiring.

 

3 hours ago, ich777 said:

Is this the first time you are trying to utilise your Nvidia GPU in a Docker container or was it working previously?

 

It was never working previously. I've been using Plex on Ubuntu for years. Got a free GTX1060 and decided I'd give it a try and see if I could get it working. Had similar issues in Ubuntu and I decided to wipe everything and start over (I also had other issues, but Ubuntu hadn't been updated since 16.04, so I figured that was the majority of the problem).

3 hours ago, ich777 said:

Please never runs such a container as Privileged, this is not a switch that will solve all your issues, exactly to other side is happening in your case because you are running the container in Host networking mode you basically give the container full access to your Hosts network interfaces and Privileged is always a thing that should be avoided at all cost.

 

Thank you! I am new to UnRAID / Docker, I have removed the switch.

 

3 hours ago, ich777 said:

Seems like it can't utilise your Nvidia GPU, please upgrade to the latest Nvidia Driver in the plugin, reboot and try to official Plex container and see if it's working with that one.

Maybe you have to install a docker mod into the Linux Server container but I'm not sure about that because I'm not familiar with their container.

OK, the only reason I moved to it is because people have said it worked for this. I am going to go back to the official container, I actually would prefer that.

 

3 hours ago, ich777 said:

Also make sure to enable Above 4G Decoding or in some cases it's named Support for Huge Address Spaces in the BIOS at the PCI submenu.

 

Please also double check if you don't have any additional spaces at the end of the UUID in the Docker template.

Everything seems fine from what I see in your Diagnostics.

 

There are no spaces in the UUID or any of the other variables.

I'll take a look in the BIOS again. It might not be possible then with my hardware... You may or may not have noticed in the Diagnostics file. My Motherboard is a very old server hardware from 2000, it's got DDR2 ECC memory max 8GB, AMD Opteron 1356. I've had this server for almost 15 years now and got it for free lol. Yesterday, while at my wits end, I started to concede that it might just not be possible on this hardware but figured I'd reach out here. Though, I'm actually surprised, the Operton 1356 can actually raw dog 1080P transcodes w/o hardware support.

 

3 hours ago, ich777 said:

ATTENTION: Please try forcing a transcode with a native Plex app like for iOS or Android since it a known issue that the Plex Web Player or Windows App just shows a black screen and transcoding is not working there in combination with Nvidia.

In the example above it was plex running transcode from Google Pixel 9 Pro.

 

5 hours ago, ich777 said:

Is this the first time you are trying to utilise your Nvidia GPU in a Docker container or was it working previously?

 

Please never runs such a container as Privileged, this is not a switch that will solve all your issues, exactly to other side is happening in your case because you are running the container in Host networking mode you basically give the container full access to your Hosts network interfaces and Privileged is always a thing that should be avoided at all cost.

 

Seems like it can't utilise your Nvidia GPU, please upgrade to the latest Nvidia Driver in the plugin, reboot and try to official Plex container and see if it's working with that one.

Maybe you have to install a docker mod into the Linux Server container but I'm not sure about that because I'm not familiar with their container.

 

Also make sure to enable Above 4G Decoding or in some cases it's named Support for Huge Address Spaces in the BIOS at the PCI submenu.

 

Please also double check if you don't have any additional spaces at the end of the UUID in the Docker template.

Everything seems fine from what I see in your Diagnostics.

 

ATTENTION: Please try forcing a transcode with a native Plex app like for iOS or Android since it a known issue that the Plex Web Player or Windows App just shows a black screen and transcoding is not working there in combination with Nvidia.

 

@ich777 got an update to this. It's still throwing the same errors. I did the following:

1. Removed plex (linuxserver) and deleted the .../appdata/plex/ folder
2. Updated video driver to v570 (current version branch)
3. Restarted PC and checked to make sure the driver installed correctly
4. Installed the official plex container (followed the guide in post 2 of this thread)

Still got the same errors. I looked through the BIOS settings, not much for PCIe settings, just Enable or Disable power control. It was disabled, I tried enabling it.

 

However, I had noticed that when using lspci, there is an integrated GPU (AMD ES1000) on the motherboard. Is it possible that somehow Plex is trying to use this GPU? If so, I'm not sure how to check or how to stop it. I have attached updated Diagnostics file.

 

nas-1-diagnostics-20250323-1305.zip

I recently upgraded the GPU from a 1660Ti that was working great for transcoding in plex, to a 1080Ti.

 

I shutdown the server, swapped GPU, and now my plex docker has disappeared and the GPU is not showing in the Nvidia Driver plugin, and the GPU Dashboard isn't showing it either.  It IS showing in the System Devices, and I can select it for the GPU stats dashboard but the dashboard shows this error:

I am also getting these errors trying to shutdown the server to pull diagnostics.

 

Diagnostics attached.  Hopefully they can help shed some light.  I put back in my old 1660Ti and everything is working as it should.

 

 

 

 

milkyway-diagnostics-20250323-1900.zip

Edited March 24, 20251 yr by Rhuarc

6 hours ago, Rhuarc said:

I recently upgraded the GPU from a 1660Ti that was working great for transcoding in plex, to a 1080Ti.

 

your diags showing a 1060

 

0c:00.0 VGA compatible controller [0300]: NVIDIA Corporation GP106 [GeForce GTX 1060 6GB] [10de:1c03] (rev a1)
	Subsystem: eVga.com. Corp. Device [3842:6368]
	Kernel driver in use: nvidia
	Kernel modules: nvidia_drm, nvidia
0c:00.1 Audio device [0403]: NVIDIA Corporation GP106 High Definition Audio Controller [10de:10f1] (rev a1)
	Subsystem: eVga.com. Corp. Device [3842:6368]

 

and when you swap your GPU, you also changed the UID of the card in the docker template ? also chacked Plex then once it started so ots still setted to a Nvidia card.

 

seeking logs, everything is looking ok in the logs with your card

11 hours ago, strider2112 said:

However, I had noticed that when using lspci, there is an integrated GPU (AMD ES1000) on the motherboard. Is it possible that somehow Plex is trying to use this GPU?

I don't think that this is the issue here since you've explicitly set the GPU to your Nvidia GPU.

 

I'm also not sure what it could be, maybe it is related to your old hardware but I'm not sure about that.

From what I can see from your Diagnostics everything should work just fine.

 

Can you please try to set up Jellyfin and see if it is working there, just to double check...

8 hours ago, Rhuarc said:

GPU stats dashboard but the dashboard shows this error

Please note that this is related to the GPU Statistics Plugin and is possibly caused because you didn't changed the UUID in the plugin settings.

Please also make sure to increase the polling rate from the plugin to something reasonable like 5000ms (=5s) or I would even recommend setting it to something higher than 20000ms (=20s).

 

8 hours ago, Rhuarc said:

I am also getting these errors trying to shutdown the server to pull diagnostics.

Are you sure that the card is working and don't has any issues?

Please double check that Above 4G Decoding alongside with Resizable BAR Support in your BIOS is enabled.

Hi @ich777,

 

Back in Feb2025 we discussed me running into a strange "nvidia-container-cli permission denied" error, seemingly out of nowhere, after I upgraded from Unraid 6.12.14--> 7.0. For reference, our discussion is here

 

Unfortunately, this issue hasn't gone away but I'm desperate for a solution. I downgraded from Unraid 7.0.1 --> 6.12.15 but still encounter the issue. I have removed the Nvidia driver plugin, turned docker on/off, rebooted, and reinstalled the Nvidia driver plugin again but still encounter the issue.

 

My question is, how can I "start fresh"? I need some help getting back to my previous state where this permissions issue doesn't happen anymore

 

Currently, my 2 main headaches that this permissions issue causes are:

 

1) I do a weekly, overnight appdata backup w/ the associated plugin. This plugin stops each docker, backs up it's appdata folder, and then restarts the container. So my Plex container (w/ Nvidia GPU) is stopped, backed up, but then can't restart again because of this nvidia-container-cli permissions error. Recall that my only workaround for this error is to install and then uninstall any plugin (I like using the CA GUI Links plugin for this but any seem to work), and then my Plex (or other docker w/ Nvidia GPU passed through) will start back up. So I end up with many hours of downtime until I'm able to manually install/uninstall a plugin the following morning and restart my Plex docker.

 

Idea - do you know if there is a way to install a plugin that's in the app store using the command line or using a script? If I could automate installing and then uninstalling a plugin using say a userscript, I could just configure that to run after my backup process. That would effectively work around this issue for me

 

2) 2nd major headache - I have my array set to auto start on boot, but my dockers w/nvidia gpu passed through always fail to start on first system boot until I install and then uninstall any plugin. This is a significant issue since I now have to manually monitor each boot. As above, if I could automate installing and uninstalling a plugin at first array start (like with a userscript), that would be a good workaround for this issue.

 

Final question - Alternatively, I'm willing to do the extreme - can I format my Unraid USB drive, install a fresh copy of Unraid 7.0.1 to it, and essentially remake my server from scratch? I've never done that before so I'm just not sure how it would work regarding not losing my 400 TB of data. If I reassign my hard drives to the array (I don't use parity) + assign my nvme cache drives to their respective, correctly named pools, would I be able to start my array without formatting my disks and access my data?

 

I appreciate any final advice you can give on this issue

Edited March 29, 20251 yr by msalad

50 minutes ago, msalad said:

I appreciate any final advice you can give on this issue

I would recommend that you grab a new flash drive, install Unraid and register trail on it.

After doing so, if possible connect a spare HDD to the server (a extrnal usb disk would also work and in the worst case a flash drive) boot into this installation, assign only the one spare HDD and start the Array, after starting the Array set the auto start to yes in the Desk Settings, install the Nvidia Driver and possibly Jellyfin or Plex with your Nvidia card assigned to the container and autostart for the container enabled.

Restart Unraid and see if you get the same error.

 

I really don‘t know what‘s going on on your system since you seem to be the only one with that issue, it wouldn‘t be the first time that someting in a User Script or somewhere else causes this issue, I guess it could be also related to the Hardware or some setting in your BIOS.

On 3/20/2025 at 4:53 AM, ich777 said:

Sorry currently I only compile the latest driver version from the open source driver which is available at the release from the Unraid version you are running.

This might change in the future but for now it is what it is, however if you are willing to upgrade to 7.1.0-beta.1 (which is stable since I run that on my production Server where I also compile and create the plugin packages for the 3rd party driver plugin) then you would be able to download this driver.

 

Background to that is that the plugin is in it's current state not fully able to detect the correct version if there are multiple version from the OpenSource driver in the repository available.

 

I hope that somehow helps, sorry for not having better news currently.

This is currently what is listed under "latest" correct? I have Unraid 7.0.1 installed and tried to get this NVIDIA driver for my 5090 to work but the Open Source Kernel Module says "NO" and the device is not found when doing nvidia-smi. It is however shown in lspci and also shows the constant error log of text repeating:

Mar 29 22:21:50 Tower kernel: NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
Mar 29 22:21:50 Tower kernel: NVRM: installed in this system requires use of the NVIDIA open kernel modules.
Mar 29 22:21:50 Tower kernel: NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:884)
Mar 29 22:21:50 Tower kernel: NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

Do I need to upgrade to the beta to get the open source module instead for the GPU?

3 hours ago, unmannedmedia said:

Do I need to upgrade to the beta to get the open source module instead for the GPU?

Please post your Diagnostics.

11 minutes ago, ich777 said:

Please post your Diagnostics.

Here it is: 

tower-diagnostics-20250330-0157.zip.crdownload

24 minutes ago, unmannedmedia said:

Here it is: 

You have to actually select the Open Source driver in the plugin at the radio buttons, then Click Update & Download, wait for the download to finish and then reboot.

 

By default the plugin installs the closed source driver.

11 minutes ago, ich777 said:

You have to actually select the Open Source driver in the plugin at the radio buttons, then Click Update & Download, wait for the download to finish and then reboot.

 

By default the plugin installs the closed source driver.

By radio buttons, do you mean the versions below the main three types? Near available versions?

 

Edited March 30, 20251 yr by unmannedmedia

18 minutes ago, unmannedmedia said:

By radio buttons, do you mean the versions below the main three types? Near available versions?

Please look again, the blue text at Latest Versions says Open Source Driver, this is the one you need for 5000 series.

8 minutes ago, ich777 said:

Please look again, the blue text at Latest Versions says Open Source Driver, this is the one you need for 5000 series.

Due to my ignorance, I assumed this version was not up to date and did not support 5000 series GPUs yet. Thank you for the support, as always!

4 hours ago, unmannedmedia said:

Due to my ignorance, I assumed this version was not up to date and did not support 5000 series GPUs yet. Thank you for the support, as always!

If you are mostly using the card for transcoding it doesn‘t matter much if you are one version behind, I might change my toolchain so that the latest version is also built for the open source driver.

I have to think about that a bit more…

3 hours ago, ich777 said:

If you are mostly using the card for transcoding it doesn‘t matter much if you are one version behind, I might change my toolchain so that the latest version is also built for the open source driver.

I have to think about that a bit more…

So I finally updated to "Open Source Driver: v570.86.16" but it shows "NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running."

I restarted the system before installing and after. Also ensured to enable and disable Docker too. Any reason why the supported version for 50 series doesn't work?

 

tower-diagnostics-20250330-1045.zip

Edited March 30, 20251 yr by unmannedmedia