docker-compose.yml

version: '3.8'

services:
  # ============================================================================
  # Core Services
  # ============================================================================

  # Redis for distributed locks and caching
  redis:
    image: redis:7-alpine
    container_name: arbitrage-redis
    ports:
      - "6379:6379"
    volumes:
      - redis_data:/data
    command: redis-server --appendonly yes
    healthcheck:
      test: [ "CMD", "redis-cli", "ping" ]
      interval: 5s
      timeout: 3s
      retries: 5
    environment:
      - TZ=UTC

  # Ollama for local LLM inference
  ollama:
    image: ollama/ollama:latest
    container_name: arbitrage-ollama
    ports:
      - "11434:11434"
    volumes:
      - ollama_data:/root/.ollama
    environment:
      - OLLAMA_HOST=0.0.0.0:11434
    healthcheck:
      test: [ "CMD", "curl", "-f", "http://localhost:11434/api/tags" ]
      interval: 10s
      timeout: 5s
      retries: 5

  # ============================================================================
  # APM and Observability Services (Issue #42)
  # ============================================================================

  # Jaeger for distributed tracing
  jaeger:
    image: jaegertracing/all-in-one:latest
    container_name: arbitrage-jaeger
    environment:
      - COLLECTOR_OTLP_ENABLED=true
      - COLLECTOR_OTLP_HOST_PORT=0.0.0.0:4317
      - SPAN_STORAGE_TYPE=badger
      - BADGER_EPHEMERAL=false
      - BADGER_DIRECTORY_VALUE=/badger/data
      - BADGER_DIRECTORY_KEY=/badger/key
    ports:
      - "6831:6831/udp"  # Jaeger agent (compact thrift)
      - "14250:14250"    # Jaeger gRPC
      - "14268:14268"    # Jaeger HTTP collector
      - "16686:16686"    # Jaeger UI (accessible at http://localhost:16686)
      - "4317:4317"      # OTLP gRPC receiver
      - "4318:4318"      # OTLP HTTP receiver
    volumes:
      - jaeger_data:/badger
    healthcheck:
      test: [ "CMD", "curl", "-f", "http://localhost:16686/api/traces" ]
      interval: 10s
      timeout: 5s
      retries: 5
    command:
      - --sampling.strategies-file=/etc/jaeger/sampling_strategies.json
    networks:
      - arbitrage-network

  # Prometheus for metrics collection
  prometheus:
    image: prom/prometheus:latest
    container_name: arbitrage-prometheus
    ports:
      - "9090:9090"
    volumes:
      - ./docker/prometheus.yml:/etc/prometheus/prometheus.yml
      - prometheus_data:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--storage.tsdb.retention.time=30d'
      - '--web.console.libraries=/usr/share/prometheus/console_libraries'
      - '--web.console.templates=/usr/share/prometheus/consoles'
    environment:
      - TZ=UTC
    healthcheck:
      test: [ "CMD", "wget", "--quiet", "--tries=1", "--spider", "http://localhost:9090/-/healthy" ]
      interval: 10s
      timeout: 5s
      retries: 5
    networks:
      - arbitrage-network

  # Grafana for visualization and dashboards
  grafana:
    image: grafana/grafana:latest
    container_name: arbitrage-grafana
    ports:
      - "3000:3000"
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=admin
      - GF_SECURITY_ADMIN_USER=admin
      - GF_INSTALL_PLUGINS=grafana-piechart-panel
      - GF_PATHS_PROVISIONING=/etc/grafana/provisioning
    volumes:
      - grafana_data:/var/lib/grafana
      - ./docker/grafana/provisioning:/etc/grafana/provisioning
      - ./docker/grafana/dashboards:/etc/grafana/provisioning/dashboards
    depends_on:
      - prometheus
    healthcheck:
      test: [ "CMD", "wget", "--quiet", "--tries=1", "--spider", "http://localhost:3000/api/health" ]
      interval: 10s
      timeout: 5s
      retries: 5
    networks:
      - arbitrage-network

  # OpenTelemetry Collector (optional, for advanced scenarios)
  otel-collector:
    image: otel/opentelemetry-collector-contrib:latest
    container_name: arbitrage-otel-collector
    command: [ "--config=/etc/otel-collector-config.yml" ]
    volumes:
      - ./docker/otel-collector-config.yml:/etc/otel-collector-config.yml
    ports:
      - "4317:4317"      # OTLP gRPC receiver
      - "4318:4318"      # OTLP HTTP receiver
      - "8889:8889"      # Prometheus exporter metrics
      - "8888:8888"      # Metrics listening endpoint
    depends_on:
      - jaeger
    environment:
      - TZ=UTC
    healthcheck:
      test: [ "CMD", "curl", "-f", "http://localhost:13133" ]
      interval: 10s
      timeout: 5s
      retries: 5
    networks:
      - arbitrage-network

  # ============================================================================
  # Application Container (placeholder for FastAPI)
  # ============================================================================
  # Uncomment when ready to run FastAPI in Docker
  # fastapi:
  #   build:
  #     context: .
  #     dockerfile: Dockerfile
  #   container_name: arbitrage-fastapi
  #   ports:
  #     - "8000:8000"
  #   environment:
  #     - ENVIRONMENT=production
  #     - APM_ENABLED=true
  #     - APM_BACKEND=jaeger
  #     - JAEGER_ENDPOINT=http://jaeger:14268/api/traces
  #     - DATABASE_URL=postgresql://user:password@postgres:5432/arbitrage_ai
  #     - REDIS_URL=redis://redis:6379/0
  #     - OLLAMA_URL=http://ollama:11434/v1
  #   depends_on:
  #     - redis
  #     - ollama
  #     - jaeger
  #     - prometheus
  #   networks:
  #     - arbitrage-network

volumes:
  redis_data:
  ollama_data:
  jaeger_data:
  prometheus_data:
  grafana_data:

networks:
  arbitrage-network:
    driver: bridge