Skip to main content

Docker Deployment

OpenRA-RL provides Docker images for reproducible deployment and headless RL training.

Quick Start​

# Build the image
cd docker && bash build.sh && cd ..

# Start the game server
docker compose up openra-rl

The server exposes:

  • Port 8000 — HTTP/WebSocket (OpenEnv protocol)
  • Port 9999 — gRPC (direct bridge access)

Docker Compose Services​

services:
  openra-rl:      # Game server
  agent:          # LLM agent (needs OPENROUTER_API_KEY)
  mcp-bot:        # MCP bot with planning tools

Run with an agent:

# Scripted bot (connect from host)
docker compose up openra-rl
python examples/scripted_bot.py

# LLM agent (containerized)
docker compose up openra-rl agent

# MCP bot (containerized)
docker compose up openra-rl mcp-bot

Headless Mode (Null Platform)​

For RL training, the Null Platform runs OpenRA without any GPU, display, or audio:

  • CPU usage: ~3% (vs 400% with Mesa software rendering)
  • No native dependencies: No SDL2, OpenAL, or display server needed
  • Docker-native: No Xvfb workarounds required

The Null Platform is enabled automatically in the Docker image via Game.Platform=Null.

Resource Limits​

Default Docker Compose limits:

deploy:
  resources:
    limits:
      cpus: '4'
      memory: 4G

Adjust in docker-compose.yaml based on your training setup.

Environment Variables​

VariableDefaultDescription
AI_SLOT—AI player slot configuration
BOT_TYPE—Built-in AI type
RECORD_REPLAYSfalseSave .orarep replay files
DISPLAY:99X11 display (ignored in Null Platform)
DOTNET_ROLL_FORWARDLatestMajor.NET runtime version policy

Building for Different Architectures​

The Docker image supports both amd64 and arm64:

# Build for current platform
docker build -f Dockerfile -t openra-rl .

# Multi-platform build
docker buildx build --platform linux/amd64,linux/arm64 -t openra-rl .

Note: On arm64, the Grpc.Tools protoc compiler crashes (SIGSEGV). The build uses pre-generated C# protobuf files with SKIP_PROTOC=true to work around this.

Health Check​

Verify the server is running:

curl http://localhost:8000/health