[Bug Report] Can't link the omniverse assets in training process inside docker container.

[Privilger]

Describe the bug

Following the docker deployment tutorial, I was able to run isaacsim inside the docker container with headless mode on a remote server workstation computer.

However, when I try to train a G1 walking policy using the following command:
/isaac-sim/python.sh source/standalone/workflows/rsl_rl/train.py --task Isaac-Velocity-Flat-G1-v0 --num_env 4096 --headless

I receive this warning and the python program is terminated.

Steps to reproduce

1. Enter the docker container with this tutorial

2. run command:
   /isaac-sim/python.sh source/standalone/workflows/rsl_rl/train.py --task Isaac-Velocity-Flat-G1-v0 --num_env 4096 --headless

3. Could not open asset @http://omniverse-content-production....
   

System Info

Describe the characteristic of your environment:

• Commit: a56291d
• Isaac Sim Version: 4.2.0
• OS: Ubuntu 20.04
• GPU: RTX L20
• CUDA: 12.2
• GPU Driver: 535

Additional context

I check my remote server computer's network to see if it is network problem.
I was able to download the asset using 'wget http://omniverse-content-production....' command to download the usd asset successful.

I also try to run the training python script in my local computer with docker container, it work well.

Checklist

• [x ] I have checked that there is no similar issue in the repo (required)
• [ x] I have checked that the issue is not in running Isaac Sim itself and is related to the repo

[jsmith-bdai]

Thanks for posting - this is strange. Can you attach a full stack trace please?

[xiaojuntime]

On host side

1. Determine local assets directory by $ docker inspect isaac-lab-base
   For instance, using following volume mapping
   host: "/var/lib/docker/volumes/docker_isaac-lab-data/_data"
   container: "/workspace/isaaclab/data_storage"
   Then create the local directory structure
   $ sudo mkdir -p /var/lib/docker/volumes/docker_isaac-lab-data/_data/Assets/Isaac/4.2/

2. Download assets to local directory
   You can pre-copy/download any assets you want to the local directory.
   For 'Isaac-Velocity-Flat-G1-v0' task, following assets are needed.
   

In isaac-lab-base container

Change the assets_path in /workspace/isaaclab/source/extensions/omni.isaac.lab/omni/isaac/lab/app/app_launcher.py from external website to local asset directory.
#assets_path = "http://omniverse-content-production.s3-us-west-2.amazonaws.com/Assets/Isaac/4.2"
assets_path = "/workspace/isaaclab/data_storage/Assets/Isaac/4.2"

Following command should be ok.
# /isaac-sim/python.sh source/standalone/workflows/rsl_rl/train.py --task Isaac-Velocity-Flat-G1-v0 --num_env 4096 --headless