Mdx-deepstream container cannot see GPUs

Accelerated Computing Intelligent Video Analytics TAO Toolkit
1

Please provide the following information when requesting support.

• Hardware dual L4 AMD64 system with
96 Gbyte RAM

I am following the setup instructions for the metropolis microservices for multicamera tracking.
(Quickstart Setup — mdx-docs 2.1 documentation)

I have confirmed that
docker run --rm --gpus all nvidia/cuda:12.2.0-base-ubuntu20.04 nvidia-smi
works.

When I bring up the other containers for the multicamera tracking, i.e. run

docker compose -f foundational/mdx-foundational.yml -f mtmc-app/mdx-mtmc-app.yml --profile e2e up -d --pull always --build --force-recreate

it brings up all containers, except for mdx-deepstream.

The docker log for the mdx-deepstream is attached:
mdx-deepstream-log.txt (3.5 KB)

It would appear the container cannot see the GPUs?

I have no idea why, since docker run --rm --gpus all nvidia/cuda:12.2.0-base-ubuntu20.04 nvidia-smi does run (albeit only after setting no-cgroups = false in /etc/nvidia-container-runtime/config.toml).

2

Seems that it is a topic for metropolis microservices. You can ask in its forum https://forums.developer.nvidia.com/c/accelerated-computing/intelligent-video-analytics/mm-ea/500 for further checking. Thanks.