Plex NVIDIA Gpu Passthrough SCALE 21.02

ornias

Wizard
Joined
Mar 6, 2020
Messages
1,458
Then you need to figure out what signals you are sending off by going "Guess you are doing it wrong then *winkyface*"
I'm not interested in your communication advice.

iX probably wants bug reports that are actually bugs.
If something doesn't work as you might expect it, that is worth a bug report. At the very least it needs to be documented appropiately.
 

BetYourBottom

Contributor
Joined
Nov 26, 2016
Messages
141
I'm not interested in your communication advice.
Then don't make incorrect claims about your condescension.

I'm sorry for getting so pissy about this but you aren't the first more experienced member of a tech website that was condescending in this same manner lately and that sort of thing grinds me.
If something doesn't work as you might expect it, that is worth a bug report. At the very least it needs to be documented appropiately.
I will consider it.
 

DaSnipe

Explorer
Joined
Jun 26, 2021
Messages
86
I have a old nvidia 740 GT that I'm also unable to passthrough to Jellyfin using TrueCharts


I've read up a flag that we need to pass through to disable health checks since our cards are pretty old and not supported by current drivers.
 

ornias

Wizard
Joined
Mar 6, 2020
Messages
1,458
@DaSnipe Thats worth creating a github bug report with, normally I would say a suggestion but as users would expect those cards to work I think a bug ("not working as expected") with link the the NVIDIA info and/or link to how that healthcheck could be disabled might be worthwhile.

As disabling that healthcheck could be automated in the middleware just fine, which would fix these issues :)
 

ornias

Wizard
Joined
Mar 6, 2020
Messages
1,458
Then don't make incorrect claims about your condescension.

Viewing your own interpentation as a fact/proof to disprove the intent of the sender and claim the sender is incorrect about his/her own intent such a big retorical mistake I'm not even going to bother with your comments about this anymore.

I'll save you from my responses in the future, welcome to my ignore list.

If I cannot even give a single plain oneliner pointing towards the bug system without being called out for all sorts of things, I do not welcome your interaction in the future. As everyone here knows I'm quite the snark, but this wasn't it.
 

BetYourBottom

Contributor
Joined
Nov 26, 2016
Messages
141
I've read up a flag that we need to pass through to disable health checks since our cards are pretty old and not supported by current drivers.
I've read similar. I'd like to know how to set it up for the underlying system TrueNAS uses because I think the likelihood of any changes to the system/UI for what almost amounts to a hack is low.
 

BetYourBottom

Contributor
Joined
Nov 26, 2016
Messages
141
Viewing your own interpentation as a fact/proof to disprove the intent of the sender and claim the sender is incorrect about his/her own intent such a big retorical mistake I'm not even going to bother with your comments about this anymore.

I'll save you from my responses in the future, welcome to my ignore list.

If I cannot even give a single plain oneliner pointing towards the bug system without being called out for all sorts of things, I do not welcome your interaction in the future. As everyone here knows I'm quite the snark, but this wasn't it.

I'm sorry for treating you so harshly. Like I failed to properly apologize for before, I got riled up because of some previous frustration with a user of a different site for being unhelpful and I wrongfully took it out on you.

I still feel like your comment was condescending and could've been expressed better. However, it didn't rise to a level that required the response I gave you and I won't argue about it further.

I'm also not going to set you to ignore in retaliation because I want to leave open the chance that I'll be able to learn something from you in the future.

I know it's unlikely that you'll read this since you set me to ignore but I thought I'd leave this here to give you a public apology.
 
Last edited:

crkinard

Explorer
Joined
Oct 24, 2019
Messages
80
Witch docker image are you using?

I was messing with the official Plex one and had some issues. Switched to the linuxserver.io one and followed their instructions and had it running inside 5 min.


You do not need the nvidia docker container as Scale already has all that. Just add
Code:
NVIDIA_VISIBLE_DEVICES=all
to the environment variables and check off use HW transcoding in the Plex settings once the image is running.

Using a Quadro K2200 (have a crap ton of these from work) and just now upgraded it to a P2000 so it has HVEC support.
Code:
root@dock[~]# nvidia-smi
Sun Nov 21 21:32:43 2021       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.91.03    Driver Version: 460.91.03    CUDA Version: 11.2     |
|-------------------------------+----------------------+----------------------+
| 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  Quadro K2200        Off  | 00000000:03:00.0 Off |                  N/A |
| 42%   34C    P8     1W /  39W |      3MiB /  4043MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
|   1  Quadro P2000        Off  | 00000000:04:00.0 Off |                  N/A |
| 45%   35C    P0    33W /  75W |    217MiB /  5059MiB |     17%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                              
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    1   N/A  N/A     68190      C   ...diaserver/Plex Transcoder      214MiB |
+-----------------------------------------------------------------------------+
 
Last edited:

ericng

Cadet
Joined
Jul 27, 2023
Messages
2
你使用的是女巫泊坞窗图像吗?

我在搞乱官方的 Plex 并遇到了一些问题。切换到 linuxserver.io 并按照他们的说明进行操作,并在 5 分钟内运行了它。


您不需要 nvidia docker 容器,因为 Scale 已经具备了所有这些。只需将
Code:
NVIDIA_VISIBLE_DEVICES=all
添加到环境变量中,并在映像运行后在 Plex 设置中勾选“使用硬件转码”即可。

使用 Quadro K2200(工作中拥有大量此类设备),现在刚刚将其升级到 P2000,因此它具有 HVEC 支持。
[代码]root@dock[~]# nvidia-smi
2021 年 11 月 21 日星期日 21:32:43
+------------------------------------------------ ----------------------------+
| NVIDIA-SMI 460.91.03 驱动程序版本:460.91.03 CUDA 版本:11.2 |
|------------------------------------------+----------------- ---+----------------------+
| GPU 名称持久性-M| 总线 ID Disp.A | 挥发性未校正。ECC |
| 风扇温度性能功率:使用/上限| 内存使用情况 | GPU-Util 计算 M。
| | | 我 M. |
|================================+================== ====+======================|
| 0 Quadro K2200 关闭 | 00000000:03:00.0 关闭 | 不适用 |
| 42% 34C P8 1W / 39W | 3MiB / 4043MiB | 0% 默认 |
| | | 不适用 |
+--------------------------------------------+----------------- ---+----------------------+
| 1 Quadro P2000 关闭 | 00000000:04:00.0 关闭 | 不适用 |
| 45% 35C P0 33W / 75W | 45% 35C P0 33W / 75W 217MiB / 5059MiB | 17% 默认 |
| | | 不适用 |
+--------------------------------------------+----------------- ---+----------------------+

+------------------------------------------------ ----------------------------+
| 流程:|
| GPU GI CI PID 类型 进程名称 GPU 内存 |
| ID ID 用途 |
|=================================================== ===========================|
| 1 不适用 不适用 68190 C ...diaserver/Plex 转码器 214MiB |
+------------------------------------------------ ----------------------------+[/代码]
Hi, I know this is a very old threat but, would you like to tell me how to add NVIDIA_VISIBLE_DEVICES=ALL to docke env??
 
Top