Error Running DLI Course Environment Container on Jetson Nano Orin with JetPack 5.11

Issue Overview

Users are encountering difficulties when attempting to run the NVIDIA Deep Learning Institute (DLI) "Getting Started with AI on Jetson Nano" course environment container on the Jetson Nano Orin board with JetPack 5.11. The specific problem arises when trying to execute the hello_camera application, resulting in an import error related to GLIBC version compatibility. This issue prevents users from proceeding with the course materials and exercises on their Jetson Nano Orin devices.

The error message indicates a mismatch between the required GLIBC version (2.29) and the available version on the system:

ImportError: /lib/aarch64-linux-gnu/libm.so.6: version `GLIBC_2.29' not found (required by /usr/lib/aarch64-linux-gnu/libgstreamer-1.0.so.0)

This problem occurs because the course container (v2.0.2-r32.7.1) is designed for JetPack 4.x and the original Jetson Nano, while the user is running JetPack 5.11 on a Jetson Nano Orin board.

Possible Causes

1. Container Incompatibility: The DLI course container is not compatible with JetPack 5.x versions, as it was designed for JetPack 4.x and the original Jetson Nano hardware.

2. GLIBC Version Mismatch: The container requires GLIBC 2.29, which is not available in the JetPack 5.11 environment.

3. Hardware Differences: The course materials and container were developed for the original Jetson Nano, not the newer Jetson Nano Orin, which may have different hardware specifications and requirements.

4. Outdated Course Materials: The DLI course has not been updated to support the latest JetPack versions and newer Jetson hardware.

Troubleshooting Steps, Solutions & Fixes

1. Use Alternative Container:

   • Instead of using the DLI course container, use the l4t-ml container version 28, which is compatible with JetPack 5.
   • Download the l4t-ml container from the NVIDIA NGC catalog.
   • Run the container using the appropriate Docker commands for your system.

2. Extract Course Materials:

   • Copy the notebooks from the original DLI container to your local system.
   • Run these notebooks within the l4t-ml container environment.

3. Install Additional Dependencies:

   • You may need to install the jetcam package within the l4t-ml container.
   • Use the following command to install jetcam:

     pip install jetcam
     

4. Explore Alternative Tutorials:

   • Consider following the "Hello AI World" tutorial, which is compatible with JetPack 5.
   • Visit the NVIDIA Developer website to find updated tutorials and guides for the Jetson Nano Orin.

5. Check for Updates:

   • Regularly check the NVIDIA DLI website and forums for announcements about course updates for newer Jetson hardware and JetPack versions.

6. Community Support:

   • Engage with the NVIDIA Developer community forums to seek additional assistance or alternative solutions.
   • Share your experience and workarounds to help other users facing similar issues.

7. Downgrade JetPack (Not Recommended):

   • As a last resort, you could consider downgrading to JetPack 4.x to use the original container. However, this is not recommended as it may introduce compatibility issues with the Jetson Nano Orin hardware.

8. Report the Issue:

   • Contact NVIDIA support or use their feedback channels to report the compatibility issue and request an update for the DLI course materials to support newer Jetson hardware and JetPack versions.

Please note that as of the last update to this discussion, there was no official announcement regarding the update of the DLI course for the Jetson Nano Orin. Users are encouraged to utilize the alternative resources mentioned above while waiting for potential future updates to the course materials.