ich777

Thank you, this is a known issue with newer drivers and I haven’t got time yet to investigate but I don‘t think that this PR will be accepted even if it will be it will break compatibility with older drivers or even the legacy drivers.

This is not the correct support thread, please click on the container Icon on the Docker page and select Support. Please post your issue there. Also a bit more information would be helpful, what settings do you use in the template? On what Unraid version are you?

Please uninstall the plugin and then reinstall it and see if that helps.

Why are you using the v470.x driver? You will be fine with the latest version anyways, you won't have this issue next time if you stay on any of the main branches from the "latest" section. Please post your Diagnostics if you have any further issues. Do you have pialert installed? They have a bug in their software which eats up all your GitHub API (which I make heavily use in my Plugins/Containers) and prevent the download. Do you have any AdBlocking on your network? Please make sure that you don't block github.com and api.github.com

@TheUseless also read this and the following post:

No. Please do the following: Reboot Go to the plugin page and install the v470.x driver version Reboot again After that it is working again. What card are you using?

This is not possible with Nvidia cards since consumer cards are limited to 3 simultaneous transcodes. However if you have many 4k content then 4GB seems the better choice to me. You have to step it up a bit if you want more transcodes to a P2000 to get unlimited transcodes, I can‘t recommend a more recent card since I‘m not familiar with newer cards which have unlimited transcodes enabled.

In case you didn't noticed, it's already in the CA App

My advice would be that you buy something faily recent and low power like a NVIDIA T400, you can get it for about 100,- and it can only consume a maximum of 35W, in idle my card pulls between 2 and 5W. But do your research first, it can do three simultaneous transcodes and has only 2GB of VRAM but that is enough for most people.

Thank you for the report. Please update the container itself by clicking "force update" (with the Advanced View enabled in the top right corner, don't forget it to disable it again after updating it), the container is now fixed and is working again.

Yes, because you are using CPU/Software transcoding which will use a lot of power and CPU resources No, hardware transcoding is not working because you are using CPU/Software transcoding and that is why your CPU is spiking. intel_gpu_top can't show anything because a ASPEED chip is not a Intel iGPU, so to speak it is not possible for intel_gpu_top any stats from an ASPEED GPU.

I don't see a iGPU in your lspci output, are you sure that you've enabled it in the BIOS or at least set it to the primary GPU? This is the only GPU I see in your system: 07:00.0 VGA compatible controller [0300]: ASPEED Technology, Inc. ASPEED Graphics Family [1a03:2000] (rev 41) DeviceName: ASPEED Video AST2500 Subsystem: Super Micro Computer Inc ASPEED Graphics Family [15d9:1a1f] Kernel driver in use: ast Kernel modules: ast

Yes, please post your Diagnostics.

Just to let you know, everything is working as expected, was a simple configuration issue.

Sorry but I only compile the following branches for my plugin: Production Branch Next Feature Branch Legacy Branch (Driver v470.x) Open Source Driver However if you are using Stable diffusion, I assume you are using it in a Docker container which runs it's own version from the Cuda Toolkit if I'm not mistaken, usually the output that you are seeing is that the driver supports up to Cuda 12.2 and should be backwards compatible with older Cuda Toolkit versions.

I gave you the wrong links, sorry, these are not for 6.12.3, here are the correct ones: https://github.com/ich777/unraid-nvidia-driver/releases/download/6.1.38-Unraid/nvidia-535.86.05-6.1.38-Unraid-1.txz https://github.com/ich777/unraid-nvidia-driver/releases/download/6.1.38-Unraid/nvidia-535.86.05-6.1.38-Unraid-1.txz.md5 Sorry, this was done by accident. You have to put them in: /boot/config/plugins/nvidia-driver/packages/6.1.38/ After that reboot.

I really can't tell what's going on over there on your side. Can you download this file? https://github.com/ich777/unraid-nvidia-driver/releases/download/6.3.13-Unraid/nvidia-535.86.05-6.3.13-Unraid-1.txz and also this: https://github.com/ich777/unraid-nvidia-driver/releases/download/6.3.13-Unraid/nvidia-535.86.05-6.3.13-Unraid-1.txz.md5 Please don't delete those files since you will need it.

This is not a very good test since this file is only about 12K in size. You should rather download something bigger: wget https://github.com/jellyfin/jellyfin/archive/refs/tags/v10.8.10.zip (but that is also a pretty bad test since this file is also only about 3M in size) What download speeds do you get with the command that I've sent you?

At that topic, I'm curious what download speeds do you get and how long does it take on your system?

Please execute this command from a Unraid terminal: wget -qO- https://raw.githubusercontent.com/ich777/unraid-plugin_update_helper/master/force_package_update | bash This will force a download from all third party plugin packages which are supported by the Plugin Update Helper and you should also see the progress bar and at which speed it is downloading, please let this process finish, it will also show you when it's safe to reboot or better speaking should look like this: root@Test:~# wget -qO- https://raw.githubusercontent.com/ich777/unraid-plugin_update_helper/master/force_package_update | bash Please don't close this window until it says finished! /boot/config/plugins/nvidia-driver/p 100%[====================================================================>] 206.36M 24.5MB/s in 8.2s Finished, please take a look at your unRAID notifications if all packages where downloaded correctly! root@Test:~# It will take some time after the download finished because it also pulls the md5 sum and compares the downloaded file, this is exactly the same what the Update & Download button does. I also can't imagine that this would make a difference if you are using the same Gateway/Internet connection.

Can you please give me a bit? I'm working on a solution for you. I can't display that in the webgui, that is simply not possible.

You have to wait until it says it is finished, like the message tells you... Some users especially from China have really slow download speeds from GitHub (where the packages are located) and this can take some time. From the Diagnostics that you've uploaded I see that you only have around 88MB from 210MB downloaded.

When you upgrade you should get a Unraid notification which says you should wait until you get another notification that it says it is safe to reboot, in your case the download didn't finish, even the upgrade window should have told you that it found third party packages and you should wait until it says it is safe to reboot. However, to solve this, please click Update & Download this should re download the driver and the plugin should then tell you to reboot if the download was successful.

Can you connect through SSH to the server when the server locks up? If yes please run: diagnostics When does this error happen exactly? Is this only happening when your server does transcode? Do you run a Plex container if yes from which repository? If you are running the official one can you try to add: --no-healthcheck to the Extra Parameters from the container (you have to enable Advanced View in the top right corner to see this line in the Docker template). Please also post your docker run command.

What is Fileflows? Is this also a Docker container? If yes it is working you say? Can you please post your docker run command?