Running Deepstream in Kubernetes

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)
dGPU
• DeepStream Version
6.0
• TensorRT Version
N/A
• NVIDIA GPU Driver Version (valid for GPU only)
515.65.01
• Issue Type( questions, new requirements, bugs)
Question

The instructions for running Deepstream containers in Docker goes something like this:

docker run --gpus all -it --rm --net=host --privileged -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -w /opt/nvidia/deepstream/deepstream-6.1 nvcr.io/nvidia/deepstream:6.1-devel

Are there any instructions available for how to run this in Kubernetes, preferably via Helm? I know passing in things like volume mounts is pretty straight forward in Helm, but at least one of the docker run options doesn’t seem to be, namely --gpus which comes from an Nvidia package. (I’m pretty sure “privileged” and “net” translate fairly well). That said, I’d much prefer to use a more “official” Helm chart if one exists.

Thanks in advance.

3

To run the command you mentioned, you must make sure the “nvidia-docker” is installed correctly. NVIDIA/nvidia-docker: Build and run Docker containers leveraging NVIDIA GPUs (github.com)

4

Well, yeah, that’s what’s necessary to run the Docker command; but what about when running the container in a Kubernetes environment via Helm?

And is there an official Helm chart for running such a container?

5

You may raise topic in docker forum. Latest NGC GPU Cloud/Docker and NVIDIA Docker topics - NVIDIA Developer Forums