Skip to content

DOCKER_SETUP.md

Latest commit

 

History

History
130 lines (96 loc) · 2.95 KB

DOCKER_SETUP.md

File metadata and controls

130 lines (96 loc) · 2.95 KB

Docker Multi-Platform Setup

This document explains how to set up Docker builds and deployment for CodeRunner across multiple platforms.

Required Secrets for GitHub Actions

To enable automatic Docker Hub publishing, add these secrets to your GitHub repository:

Setting up Docker Hub Secrets

  1. Go to your Docker Hub account:

    • Visit hub.docker.com
    • Go to Account Settings → Security → New Access Token
    • Create a token with Read, Write, Delete permissions

  2. Add secrets to GitHub repository:

    • Go to your GitHub repo → Settings → Secrets and variables → Actions
    • Add these repository secrets:
DOCKER_USERNAME: your-dockerhub-username
DOCKER_TOKEN: your-dockerhub-access-token

Workflow Triggers

The Docker build workflow runs on:

  • Push to main/develop: Builds and pushes docker-latest tag
  • Pull Requests: Builds only (no push) for testing
  • Releases: Builds and pushes version tags like v1.0.0-docker

Docker Usage

Quick Start with Docker

# Run CodeRunner with Docker
docker run -d \
  --name coderunner \
  -p 8222:8222 \
  -v ./uploads:/app/uploads \
  instavm/coderunner:docker-latest

# Access MCP server
curl http://localhost:8222/mcp

Docker Compose (Recommended)

# Use the provided docker-compose.yml
docker compose up -d

# View logs
docker compose logs -f

# Stop service
docker compose down

Environment Variables

  • FASTMCP_HOST: Server host (default: 0.0.0.0)
  • FASTMCP_PORT: Server port (default: 8222)

Platform Support

Platform Container Tool Image Tag Registry
macOS (Apple Silicon) Apple container apple-latest OCI-compatible
Linux (x64/ARM64) Docker docker-latest Docker Hub
Windows Docker Desktop docker-latest Docker Hub

Integration Examples

With Claude Desktop (Docker)

{
  "mcpServers": {
    "coderunner": {
      "command": "docker",
      "args": [
        "run", "--rm", "-i",
        "-p", "8222:8222",
        "-v", "./uploads:/app/uploads",
        "instavm/coderunner:docker-latest"
      ]
    }
  }
}

With OpenAI Agents

# Update openai_client.py to use Docker endpoint
hostname = "localhost"  # Instead of coderunner.local
address = "127.0.0.1"
url = f"http://{address}:8222/mcp"

Building Locally

# Standard Docker build
docker build -t coderunner-local .

# Multi-architecture build
docker buildx build --platform linux/amd64,linux/arm64 -t coderunner-multiarch .

Troubleshooting

Common Issues

  1. Port conflicts: Change port mapping -p 8223:8222
  2. Volume permissions: Ensure ./uploads directory exists and is writable
  3. Memory issues: Add memory limits --memory 4g

Health Checks

# Check container status
docker ps

# Check logs
docker logs coderunner

# Test MCP endpoint
curl -f http://localhost:8222/mcp || echo "MCP not ready"