docker-compose.yml

# HERALD docker-compose
#
# All containers are prefixed `herald_` and attach to the external network
# named by HERALD_NETWORK_NAME. Each container is assigned a static IPv4 on
# that network via env vars (HERALD_*_IP) so addressing is stable across
# redeploys.
#
# Bundled databases (postgres, redis, timescaledb) live under the `bundled-db`
# profile so they only start when COMPOSE_PROFILES contains that name.
#   Standalone:    COMPOSE_PROFILES=bundled-db
#   External DBs:  COMPOSE_PROFILES=              (and point *_HOST at them)

services:
  # ── infrastructure (bundled) ────────────────────────────────────────────────
  postgres:
    image: postgres:16-alpine
    container_name: herald_postgres
    profiles: ["bundled-db"]
    environment:
      POSTGRES_USER: ${BOOTSTRAP_PG_SUPERUSER:-postgres}
      POSTGRES_PASSWORD: ${BOOTSTRAP_PG_SUPERPASS}
      POSTGRES_DB: postgres
    volumes:
      - postgres_data:/var/lib/postgresql/data
    ports:
      - "5432:5432"
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ${BOOTSTRAP_PG_SUPERUSER:-postgres}"]
      interval: 10s
      timeout: 5s
      retries: 10
    networks:
      herald-net:
        ipv4_address: ${HERALD_POSTGRES_IP}
    restart: unless-stopped

  timescaledb:
    image: timescale/timescaledb:latest-pg16
    container_name: herald_timescaledb
    profiles: ["bundled-db"]
    environment:
      POSTGRES_USER: ${BOOTSTRAP_TS_SUPERUSER:-postgres}
      POSTGRES_PASSWORD: ${BOOTSTRAP_TS_SUPERPASS}
      POSTGRES_DB: postgres
    volumes:
      - timescale_data:/var/lib/postgresql/data
    ports:
      - "5433:5432"
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ${BOOTSTRAP_TS_SUPERUSER:-postgres}"]
      interval: 10s
      timeout: 5s
      retries: 10
    networks:
      herald-net:
        ipv4_address: ${HERALD_TIMESCALE_IP}
    restart: unless-stopped

  redis:
    image: redis:7-alpine
    container_name: herald_redis
    profiles: ["bundled-db"]
    command: ["redis-server", "--save", "60", "1", "--loglevel", "warning"]
    volumes:
      - redis_data:/data
    ports:
      - "6379:6379"
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 10s
      timeout: 5s
      retries: 10
    networks:
      herald-net:
        ipv4_address: ${HERALD_REDIS_IP}
    restart: unless-stopped

  # ── RSSHub + Browserless (third-party images) ──────────────────────────────
  rsshub:
    image: diygod/rsshub:latest
    container_name: herald_rsshub
    environment:
      NODE_ENV: production
      CACHE_TYPE: redis
      REDIS_URL: redis://${REDIS_HOST}:${REDIS_PORT}/${REDIS_DB_RSSHUB:-1}
      PUPPETEER_WS_ENDPOINT: ws://${HERALD_BROWSERLESS_IP}:3000
      REQUEST_TIMEOUT: "30000"
    networks:
      herald-net:
        ipv4_address: ${HERALD_RSSHUB_IP}
    restart: unless-stopped

  browserless:
    image: ghcr.io/browserless/chromium:latest
    container_name: herald_browserless
    environment:
      CONCURRENT: "5"
      TIMEOUT: "30000"
      TOKEN: ""
    networks:
      herald-net:
        ipv4_address: ${HERALD_BROWSERLESS_IP}
    restart: unless-stopped

  # ── application services ────────────────────────────────────────────────────
  feed_poller:
    build:
      context: .
      dockerfile: services/feed_poller/Dockerfile
    container_name: herald_feed_poller
    env_file: .env
    environment:
      SOURCES_FILE: /app/sources/feeds.yaml
    networks:
      herald-net:
        ipv4_address: ${HERALD_FEED_POLLER_IP}
    restart: unless-stopped

  # NOTE: the extractor is single-container here because Compose forbids
  # `container_name` + `deploy.replicas`. To horizontally scale extraction,
  # launch additional containers with distinct names (e.g. herald_extractor_2)
  # from a sidecar compose file or `docker run`; they'll join the same Redis
  # consumer group automatically via EXTRACTOR_CONSUMER_GROUP.
  extractor:
    build:
      context: .
      dockerfile: services/extractor/Dockerfile
    container_name: herald_extractor
    env_file: .env
    networks:
      herald-net:
        ipv4_address: ${HERALD_EXTRACTOR_IP}
    restart: unless-stopped

  gdelt_ingestor:
    build:
      context: .
      dockerfile: services/gdelt_ingestor/Dockerfile
    container_name: herald_gdelt_ingestor
    env_file: .env
    networks:
      herald-net:
        ipv4_address: ${HERALD_GDELT_INGESTOR_IP}
    restart: unless-stopped

  maintenance:
    build:
      context: .
      dockerfile: services/maintenance/Dockerfile
    container_name: herald_maintenance
    env_file: .env
    networks:
      herald-net:
        ipv4_address: ${HERALD_MAINTENANCE_IP}
    restart: unless-stopped

  # ── one-off backfill (docker compose --profile backfill run --rm backfill ...) ─
  backfill:
    build:
      context: .
      dockerfile: services/backfill/Dockerfile
    image: herald_backfill
    container_name: herald_backfill
    profiles: ["backfill"]
    env_file: .env
    networks:
      herald-net:
        ipv4_address: ${HERALD_BACKFILL_IP}
    restart: "no"

  mcp_server:
    build:
      context: .
      dockerfile: services/mcp_server/Dockerfile
    container_name: herald_mcp_server
    env_file: .env
    environment:
      SOURCES_FILE: /app/sources/feeds.yaml
    ports:
      - "${MCP_PORT:-8000}:8000"
    networks:
      herald-net:
        ipv4_address: ${HERALD_MCP_SERVER_IP}
    restart: unless-stopped

networks:
  herald-net:
    external: true
    name: ${HERALD_NETWORK_NAME:-herald-net}

volumes:
  postgres_data:
  timescale_data:
  redis_data: