Running Deepstream 5.1 docker on Jetpack 4.6

I am seeing some strangeness related to running older docker images on newer jetpack versions.

I have a docker image based on nvcr.io/nvidia/deepstream-l4t:5.1-21.02-samples and I want to run it on Jetpack 4.6.

When I try to run our deepstream app it it seems to want to find libnvinfer.so.7.1.3. This appears to have a zero size in the filesystem inside the docker, and does not exist at all outside on the host. The file libnvinfer.so has a simlink to libnvinfer.so.8.0.1on the host Jetpack 4.6.

Is there a dependency on Deepstream docker versions vs Jetpack 4.6 versions? I think the nvidia docker runtime does some funky things like mapping external libraries in so that might mean there is a dependency, however I think this goes against the point of docker to have all the dependencies inside?

Thanks.

1 Like

Your topic was posted in the wrong category. I am moving this to the DeepStream SDK category for visibility.

Please provide complete information as applicable to your setup.
Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

Are you kidding? All of the information is contained in the question. In case you’re not:

Hardware Platform (Jetson / GPU)

deepstream-l4t, Jetpack => Jetson

• DeepStream Version

deepstream-l4t:5.1-21.02-samples => Deepstream 5.1

• JetPack Version (valid for Jetson only)

Jetpack 4.6 => Jetpack 4.6

• TensorRT Version

Jetpack 4.6 ships with 8.0.1 - you can look that up here: https://developer.nvidia.com/jetpack-sdk-46
My Avermedia NX211B Xavier NX board has 8.0.1.6

• NVIDIA GPU Driver Version (valid for GPU only)

N/A

• Issue Type( questions, new requirements, bugs)

question

• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)

Try to run an example using the nvcr.io/nvidia/deepstream-l4t:5.1-21.02-samples container on any jetson running Jetpack 4.6

• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

Requirement: Answer the following question:

Is there a dependency on Deepstream docker versions vs Jetpack versions? Do you need to run a specific deepstream docker that matches a specific Jetpack version?

2 Likes

As per documentation docker are not compatible with each other
for deepstream app working in container, you need to mount bsp related library dependency into container from the host.
https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_docker_containers.html#what-does-this-mean-for-ds-users

1 Like

Thank you, this was the key bit of information that I could not find.

It seems that the newer images have a more traditional approach to docker with all the dependencies installed.

Thanks!

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.