GPU transcoding on Docker 19.03 with Nvidia 1050TI

Hi

Just been rebuilding my entire server stack setup after moving house last month. Current it works, but I am about to have a heap of family and friends ravage the server due to the COVID-19 Lockdown, and I noticed that the plex server isn’t hardware transcoding as it is supposed to.

Running: docker run --gpus all ubuntu:16.04 nvidia-smi
Results in:
Thu Mar 26 01:49:51 2020 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 440.64.00 Driver Version: 440.64.00 CUDA Version: N/A | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | |===============================+======================+======================| | 0 GeForce GTX 105... Off | 00000000:23:00.0 Off | N/A | | 41% 36C P0 N/A / 75W | 0MiB / 4038MiB | 0% Default | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | No running processes found | +-----------------------------------------------------------------------------+

which to me indicates that the GPU is been passed through to containers…
however running a transcode in plex, it only does it via cpu no HW transcode.
for ref: nvidia-smi run from hardware… (cuda version info not found in the container)

Thu Mar 26 01:52:15 2020 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 440.64.00 Driver Version: 440.64.00 CUDA Version: 10.2 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | |===============================+======================+======================| | 0 GeForce GTX 105... Off | 00000000:23:00.0 Off | N/A | | 41% 36C P0 N/A / 75W | 0MiB / 4038MiB | 0% Default | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | No running processes found | +-----------------------------------------------------------------------------+

Server Version#: 1.18.9.2571
Docker Version#: 19.03
GPU = NVIDA 1050TI (Modded Firmware for unlimited streams)
Base OS: Ubuntu

Docker Setup:
docker run -d --gpus all --name plex --network=pub_net --ip=192.168.1.25 -e TZ="$Timezone" -e PLEX_CLAIM="" -h $hostname -v /local1/plex/plexdata:/config -v /local1/plex/transcode:/transcode -v /mnt/media:/data plexinc/pms-docker:plexpass

Fixed it I think…

I assume it was the NVIDIA vars but yeah its now working and I’m not gonna touch it now…

docker run \
-d \
--name plex \
--network=pub_net \
--ip=192.168.1.25 \
--gpus all \
-e TZ=<timezone> \
-e PLEX_CLAIM="<claimToken>" \
-e NVIDIA_VISIBLE_DEVICES="all" \
-e NVIDIA_DRIVER_CAPABILITIES="compute,video,utility" \
-h <hostname> \
-v /local1/plex/plexdata:/config \
-v /local1/plex/transcode:/transcode \
-v /mnt/media:/data \
--device /dev/dri:/dev/dri \
plexinc/pms-docker:plexpass
2 Likes

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.