ROS 2 Iron + Gazebo Garden + CUDA in Docker (WSL2)

This repository provides a step-by-step guide to set up ROS 2 Iron Irwini with Gazebo Garden inside Docker on WSL2, with full NVIDIA GPU acceleration.

Setup Instructions

1. Install Docker Desktop & WSL2

1. Install Docker Desktop from https://www.docker.com/products/docker-desktop/.
2. Enable WSL2 backend in Docker Desktop settings.
3. Install Ubuntu 22.04 in WSL2:

   wsl --install -d Ubuntu-22.04

2. Ensure WSL and Docker Desktop Are Running

1. Open PowerShell and start WSL:

   wsl

2. Inside the WSL terminal, check that Docker is running:

   docker --version

   Expected output:

   Docker version X.XX.XX, build XXXXXXX
   

3. If Docker is not running, start Docker Desktop from Windows and retry.

3. Install NVIDIA GPU Drivers & Toolkit

Run the following inside WSL:

sudo apt update && sudo apt upgrade -y
bash scripts/install_nvidia_wsl.sh

4. Build & Run the Docker Container

cd docker
docker build -t ros2-iron-gazebo-gpu .
docker-compose up -d

5. Attach to the Running Container

docker attach ros2-gpu-container

Verification Steps

1. Verify GPU Support

Inside the container, run:

nvidia-smi
glxinfo -B

Expected output:

• nvidia-smi should detect the GPU.
• glxinfo -B should show OpenGL renderer string: NVIDIA.

2. Launch Gazebo Garden with an Empty World

Start Gazebo inside the container:

source /opt/ros/iron/setup.bash
gz sim -s /usr/share/gz/gz-sim7/worlds/empty.sdf -r -v 4 & gz sim -g &

• The Gazebo GUI should open, displaying an empty simulation world.

3. Check if Gazebo is Running on the GPU

While Gazebo is open, run:

nvidia-smi

Expected output:

• The Gazebo process should be listed under GPU usage, confirming it's utilizing the NVIDIA GPU.

Troubleshooting

Check TROUBLESHOOTING.md for common issues.

License

MIT License - see LICENSE for details.