simstudioai--sim/docker-compose.ollama.yml

name: sim-with-ollama

services:
  # Main Sim Application
  simstudio:
    build:
      context: .
      dockerfile: docker/app.Dockerfile
    ports:
      - '3000:3000'
    deploy:
      resources:
        limits:
          memory: 8G
    environment:
      - DATABASE_URL=postgresql://${POSTGRES_USER:-postgres}:${POSTGRES_PASSWORD:-postgres}@db:5432/${POSTGRES_DB:-simstudio}
      - BETTER_AUTH_URL=${NEXT_PUBLIC_APP_URL:-http://localhost:3000}
      - NEXT_PUBLIC_APP_URL=${NEXT_PUBLIC_APP_URL:-http://localhost:3000}
      - BETTER_AUTH_SECRET=${BETTER_AUTH_SECRET:-sim_auth_secret_$(openssl rand -hex 16)}
      - ENCRYPTION_KEY=${ENCRYPTION_KEY:-$(openssl rand -hex 32)}
      - COPILOT_API_KEY=${COPILOT_API_KEY}
      - MSHIP_SYSPROMPT_OVERRIDE=${MSHIP_SYSPROMPT_OVERRIDE:-}
      - NEXT_PUBLIC_CHAT_DISABLED=${NEXT_PUBLIC_CHAT_DISABLED:-}
      - SIM_AGENT_API_URL=${SIM_AGENT_API_URL}
      - OLLAMA_URL=http://ollama:11434
      - NEXT_PUBLIC_SOCKET_URL=${NEXT_PUBLIC_SOCKET_URL:-}
    depends_on:
      db:
        condition: service_healthy
      migrations:
        condition: service_completed_successfully
      realtime:
        condition: service_healthy
    healthcheck:
      test: ['CMD', 'wget', '--spider', '--quiet', 'http://127.0.0.1:3000']
      interval: 90s
      timeout: 5s
      retries: 3
      start_period: 10s
    restart: unless-stopped

  # Realtime Socket Server
  realtime:
    build:
      context: .
      dockerfile: docker/realtime.Dockerfile
    environment:
      - DATABASE_URL=postgresql://${POSTGRES_USER:-postgres}:${POSTGRES_PASSWORD:-postgres}@db:5432/${POSTGRES_DB:-simstudio}
      - NEXT_PUBLIC_APP_URL=${NEXT_PUBLIC_APP_URL:-http://localhost:3000}
      - BETTER_AUTH_URL=${BETTER_AUTH_URL:-http://localhost:3000}
      - BETTER_AUTH_SECRET=${BETTER_AUTH_SECRET:-sim_auth_secret_$(openssl rand -hex 16)}
    depends_on:
      db:
        condition: service_healthy
    restart: unless-stopped
    ports:
      - '3002:3002'
    deploy:
      resources:
        limits:
          memory: 1G
    healthcheck:
      test: ['CMD', 'wget', '--spider', '--quiet', 'http://127.0.0.1:3002/health']
      interval: 90s
      timeout: 5s
      retries: 3
      start_period: 10s

  # Database Migrations
  migrations:
    build:
      context: .
      dockerfile: docker/db.Dockerfile
    environment:
      - DATABASE_URL=postgresql://${POSTGRES_USER:-postgres}:${POSTGRES_PASSWORD:-postgres}@db:5432/${POSTGRES_DB:-simstudio}
    depends_on:
      db:
        condition: service_healthy
    command: ['bun', 'run', 'db:migrate']
    restart: 'no'

  # PostgreSQL Database with Vector Extension
  db:
    image: pgvector/pgvector:pg17
    restart: always
    ports:
      - '${POSTGRES_PORT:-5432}:5432'
    environment:
      - POSTGRES_USER=${POSTGRES_USER:-postgres}
      - POSTGRES_PASSWORD=${POSTGRES_PASSWORD:-postgres}
      - POSTGRES_DB=${POSTGRES_DB:-simstudio}
    volumes:
      - postgres_data:/var/lib/postgresql/data
    healthcheck:
      test: ['CMD-SHELL', 'pg_isready -U postgres']
      interval: 5s
      timeout: 5s
      retries: 5

  # Ollama with GPU support (default)
  ollama:
    profiles:
      - gpu
    image: ollama/ollama:latest
    pull_policy: always
    volumes:
      - ollama_data:/root/.ollama
    ports:
      - '11434:11434'
    environment:
      - NVIDIA_DRIVER_CAPABILITIES=all
      - OLLAMA_LOAD_TIMEOUT=-1
      - OLLAMA_KEEP_ALIVE=-1
      - OLLAMA_DEBUG=1
      - OLLAMA_HOST=0.0.0.0:11434
    command: 'serve'
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
    healthcheck:
      test: ['CMD', 'ollama', 'list']
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 30s
    restart: unless-stopped

  # Ollama CPU-only version (use with --profile cpu profile)
  ollama-cpu:
    profiles:
      - cpu
    image: ollama/ollama:latest
    pull_policy: always
    volumes:
      - ollama_data:/root/.ollama
    ports:
      - '11434:11434'
    environment:
      - OLLAMA_LOAD_TIMEOUT=-1
      - OLLAMA_KEEP_ALIVE=-1
      - OLLAMA_DEBUG=1
      - OLLAMA_HOST=0.0.0.0:11434
    command: 'serve'
    healthcheck:
      test: ['CMD', 'ollama', 'list']
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 30s
    restart: unless-stopped
    networks:
      default:
        aliases:
          - ollama

  # Helper container to pull models automatically
  model-setup:
    image: ollama/ollama:latest
    profiles:
      - setup
    volumes:
      - ollama_data:/root/.ollama
    environment:
      - OLLAMA_HOST=ollama:11434
    entrypoint: ["/bin/sh", "-lc"]
    command: >
      sh -c "
        echo 'Waiting for Ollama to be ready...' &&
        until ollama list >/dev/null 2>&1; do echo 'Waiting for Ollama...'; sleep 2; done &&
        echo 'Pulling gemma3:4b model (recommended starter model)...' &&
        ollama pull gemma3:4b &&
        echo 'Model setup complete! You can now use gemma3:4b in Sim.' &&
        echo 'To add more models, run: docker compose -f docker-compose.ollama.yml exec ollama ollama pull <model-name>'
      "
    restart: 'no'

volumes:
  postgres_data:
  ollama_data: