bridge/compose.mainnet.yml

services:
  # Infrastructure Services
  nats:
    image: nats:2.10-alpine
    container_name: bridge-nats
    restart: always
    ports:
      - "4222:4222"
      - "8222:8222"
    command: "-js -m 8222 --store_dir /data -c /etc/nats/nats.conf"
    volumes:
      - nats-data:/data
      - ./config/nats/mainnet.conf:/etc/nats/nats.conf:ro
    networks:
      - bridge-network
    healthcheck:
      test: ["CMD", "nc", "-z", "localhost", "4222"]
      interval: 5s
      timeout: 3s
      retries: 5

  consul:
    image: hashicorp/consul:1.17
    container_name: bridge-consul
    restart: always
    ports:
      - "8500:8500"
      - "8600:8600/udp"
    command: "agent -server -bootstrap-expect=3 -ui -client=0.0.0.0 -datacenter=mainnet"
    volumes:
      - consul-data:/consul/data
      - ./config/consul/mainnet.json:/consul/config/mainnet.json:ro
    networks:
      - bridge-network
    environment:
      - CONSUL_BIND_INTERFACE=eth0
      - CONSUL_DATACENTER=mainnet
    healthcheck:
      test: ["CMD", "consul", "members"]
      interval: 5s
      timeout: 3s
      retries: 5

  # Database with replication support
  postgres:
    image: ghcr.io/hanzoai/sql:latest
    container_name: bridge-postgres
    restart: always
    environment:
      - POSTGRES_USER=bridge
      - POSTGRES_PASSWORD=${POSTGRES_PASSWORD}
      - POSTGRES_DB=bridge
      - POSTGRES_INITDB_ARGS=--encoding=UTF-8 --locale=en_US.utf8
      - POSTGRES_HOST_AUTH_METHOD=scram-sha-256
    volumes:
      - postgres-data:/var/lib/postgresql/data
      - ./scripts/init.sql:/docker-entrypoint-initdb.d/init.sql
      - ./config/postgres/mainnet.conf:/etc/postgresql/postgresql.conf:ro
    ports:
      - "5432:5432"
    networks:
      - bridge-network
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U bridge"]
      interval: 5s
      timeout: 3s
      retries: 5

  # MPC Nodes - Distributed across different availability zones
  mpc-node-0:
    image: ghcr.io/luxfi/lux-mpc:latest
    container_name: bridge-mpc-0
    restart: always
    environment:
      - NODE_ID=mpc-node-0
      - NODE_PORT=6000
      - NATS_URL=nats://nats:4222
      - CONSUL_URL=http://consul:8500
      - DATA_DIR=/data/mpc
      - LOG_LEVEL=info
      - MPC_THRESHOLD=2
      - MPC_TOTAL_NODES=3
      - MPC_KEY_PREFIX=bridge/mpc/mainnet
      - NETWORK=mainnet
      - ENABLE_TLS=true
      - TLS_CERT_PATH=/certs/tls.crt
      - TLS_KEY_PATH=/certs/tls.key
    volumes:
      - mpc-data-0:/data/mpc
      - ./certs/mpc-0:/certs:ro
    ports:
      - "6000:6000"
      - "9090:9090"
    networks:
      - bridge-network
    depends_on:
      nats:
        condition: service_healthy
      consul:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "curl", "-f", "https://localhost:6000/health", "--insecure"]
      interval: 10s
      timeout: 5s
      retries: 3
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4G
        reservations:
          cpus: '1'
          memory: 2G

  mpc-node-1:
    image: ghcr.io/luxfi/lux-mpc:latest
    container_name: bridge-mpc-1
    restart: always
    environment:
      - NODE_ID=mpc-node-1
      - NODE_PORT=6000
      - NATS_URL=nats://nats:4222
      - CONSUL_URL=http://consul:8500
      - DATA_DIR=/data/mpc
      - LOG_LEVEL=info
      - MPC_THRESHOLD=2
      - MPC_TOTAL_NODES=3
      - MPC_KEY_PREFIX=bridge/mpc/mainnet
      - NETWORK=mainnet
      - ENABLE_TLS=true
      - TLS_CERT_PATH=/certs/tls.crt
      - TLS_KEY_PATH=/certs/tls.key
    volumes:
      - mpc-data-1:/data/mpc
      - ./certs/mpc-1:/certs:ro
    ports:
      - "6001:6000"
      - "9091:9090"
    networks:
      - bridge-network
    depends_on:
      nats:
        condition: service_healthy
      consul:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "curl", "-f", "https://localhost:6000/health", "--insecure"]
      interval: 10s
      timeout: 5s
      retries: 3
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4G
        reservations:
          cpus: '1'
          memory: 2G

  mpc-node-2:
    image: ghcr.io/luxfi/lux-mpc:latest
    container_name: bridge-mpc-2
    restart: always
    environment:
      - NODE_ID=mpc-node-2
      - NODE_PORT=6000
      - NATS_URL=nats://nats:4222
      - CONSUL_URL=http://consul:8500
      - DATA_DIR=/data/mpc
      - LOG_LEVEL=info
      - MPC_THRESHOLD=2
      - MPC_TOTAL_NODES=3
      - MPC_KEY_PREFIX=bridge/mpc/mainnet
      - NETWORK=mainnet
      - ENABLE_TLS=true
      - TLS_CERT_PATH=/certs/tls.crt
      - TLS_KEY_PATH=/certs/tls.key
    volumes:
      - mpc-data-2:/data/mpc
      - ./certs/mpc-2:/certs:ro
    ports:
      - "6002:6000"
      - "9092:9090"
    networks:
      - bridge-network
    depends_on:
      nats:
        condition: service_healthy
      consul:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "curl", "-f", "https://localhost:6000/health", "--insecure"]
      interval: 10s
      timeout: 5s
      retries: 3
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4G
        reservations:
          cpus: '1'
          memory: 2G

  # Bridge Server with HA
  bridge-server:
    image: ghcr.io/luxfi/bridge-server:latest
    container_name: bridge-server
    restart: always
    environment:
      - NODE_ENV=production
      - PORT=3000
      - DATABASE_URL=postgresql://bridge:${POSTGRES_PASSWORD}@postgres:5432/bridge?sslmode=require
      - NATS_URL=nats://nats:4222
      - CONSUL_URL=http://consul:8500
      - NODE_ID=bridge-server-0
      - MPC_THRESHOLD=2
      - MPC_TOTAL_NODES=3
      - MPC_KEY_PREFIX=bridge/mpc/mainnet
      - NETWORK=mainnet
      - ENABLE_METRICS=true
      - METRICS_PORT=9090
      - LOG_LEVEL=info
      - RATE_LIMIT_ENABLED=true
      - RATE_LIMIT_MAX=100
      - RATE_LIMIT_WINDOW=60
    ports:
      - "3000:3000"
      - "9093:9090"
    networks:
      - bridge-network
    depends_on:
      postgres:
        condition: service_healthy
      nats:
        condition: service_healthy
      consul:
        condition: service_healthy
      mpc-node-0:
        condition: service_healthy
      mpc-node-1:
        condition: service_healthy
      mpc-node-2:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:3000/health"]
      interval: 10s
      timeout: 5s
      retries: 3
    deploy:
      resources:
        limits:
          cpus: '4'
          memory: 8G
        reservations:
          cpus: '2'
          memory: 4G

  # Bridge UI - Mainnet — Vite SPA (ghcr.io/hanzoai/spa serves /dist on :80).
  # BRIDGE_* env documents what the image was built with; the values
  # become real only when baked at build time (app/bridge/Dockerfile).
  bridge-ui:
    image: ghcr.io/luxfi/bridge-ui:latest
    container_name: bridge-ui
    restart: always
    environment:
      - BRIDGE_API_HOST=https://bridge-api.lux.network
      - BRIDGE_ENV=mainnet
      - BRIDGE_CLIENT_ID=lux-bridge
      - BRIDGE_IAM_ORG=lux-network
    ports:
      - "3001:80"
    networks:
      - bridge-network
    depends_on:
      - bridge-server
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 2G
        reservations:
          cpus: '1'
          memory: 1G

  # Bridge unified Go binary (cmd/bridge). Mainnet equivalent of the
  # bridge-go service in compose.testnet.yml and the Deployment in
  # k8s/bridge-deployment.yaml — same image, same env contract, same
  # mainnet chain set (networks.mainnet.yaml mirrors the k8s ConfigMap).
  #
  # Build locally before `docker compose up bridge-go`:
  #   docker build -f cmd/bridge/Dockerfile -t ghcr.io/luxfi/bridge:dev .
  # Then export BRIDGE_TAG=dev. For production, pin to a released tag
  # (e.g. BRIDGE_TAG=v1.2.3) rather than :latest so a registry retag
  # can't roll the running stack out from under you.
  #
  # Runs alongside bridge-server + bridge-ui during the §13.6 cutover
  # soak. Once the Go binary handles real traffic without regressions,
  # DNS at bridge.lux.network repoints to this service and the legacy
  # bridge-server + bridge-ui can be retired.
  bridge-go:
    image: ghcr.io/luxfi/bridge:${BRIDGE_TAG:-latest}
    container_name: bridge-go
    restart: always
    ports:
      - "8080:8080"
    environment:
      - BRIDGE_ADDR=:8080
      # The operator listener: /metrics, the swap list, the deposit poll.
      # Deliberately absent from `ports:` — it is reachable on the compose
      # network (prometheus scrapes bridge-go:8081/metrics) and not from
      # the host.
      - BRIDGE_ADMIN_ADDR=:8081
      - BRIDGE_CONFIG=/etc/bridge/networks.yaml
      - BRIDGE_DATA_DIR=/var/lib/lux-bridge
      # Empty disables the legacy backend proxy — all /v1/bridge/*
      # routes resolve through native handlers.
      - BRIDGE_BACKEND_URL=
      # BridgeVM (b-chain) JSON-RPC. Leave unset until BridgeVM exposes
      # its public API; native handlers serve quote + swap state in the
      # meantime.
      - BRIDGE_BCHAIN_URL=
      # MPC daemon. The lux-mpc image exposes its HTTP API on NODE_PORT
      # (default 6000 per the in-compose mpc-node-* env). Override
      # BRIDGE_MPC_URL to point at an external mpcd cluster when not
      # running the in-compose MPC services.
      - BRIDGE_MPC_URL=${BRIDGE_MPC_URL:-http://mpc-node-0:6000}
      - BRIDGE_MPC_TOKEN=${BRIDGE_MPC_TOKEN:-}
      - BRIDGE_MPC_ORG_ID=bridge
      # Mainnet source-chain RPC overrides. Public RPC endpoints are
      # fine for low-volume soak; pin to a private provider (Alchemy /
      # Infura / Quicknode) before high-volume cutover so rate limits
      # don't drop deposit-watcher polls.
      - BRIDGE_SOURCE_RPC_OVERRIDES=ETHEREUM_MAINNET=https://eth.llamarpc.com,LUX_MAINNET=https://api.lux.network/v1/chain/C/rpc,ARBITRUM_MAINNET=https://arb1.arbitrum.io/rpc,OPTIMISM_MAINNET=https://mainnet.optimism.io,BASE_MAINNET=https://mainnet.base.org,POLYGON_MAINNET=https://polygon-rpc.com,BSC_MAINNET=https://bsc-dataseed.binance.org,AVAX_MAINNET=https://api.avax.network/ext/bc/C/rpc
    volumes:
      # Networks + tokens config. Read-only mount so an accidental write
      # inside the container can't corrupt the operator-checked source-
      # of-truth file. Keep this in sync with the inline YAML in
      # k8s/bridge-deployment.yaml's ConfigMap.
      - ./cmd/bridge/networks.mainnet.yaml:/etc/bridge/networks.yaml:ro
      # zapdb-backed swap store + release-wallets.json. Bind-mounted to
      # a host SSD path so the data survives compose down/up and is
      # visible to host-side backup tooling.
      - bridge-go-data:/var/lib/lux-bridge
    networks:
      - bridge-network
    depends_on:
      mpc-node-0:
        condition: service_healthy
      mpc-node-1:
        condition: service_healthy
      mpc-node-2:
        condition: service_healthy
    deploy:
      resources:
        limits:
          cpus: '4'
          memory: 8G
        reservations:
          cpus: '2'
          memory: 4G

  # Enhanced Monitoring
  prometheus:
    image: prom/prometheus:latest
    container_name: bridge-prometheus
    restart: always
    volumes:
      - ./monitoring/prometheus/mainnet.yml:/etc/prometheus/prometheus.yml
      - prometheus-data:/prometheus
    ports:
      - "9094:9090"
    networks:
      - bridge-network
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--storage.tsdb.retention.time=30d'
      - '--web.enable-lifecycle'
    deploy:
      resources:
        limits:
          cpus: '1'
          memory: 2G

  grafana:
    image: grafana/grafana:latest
    container_name: bridge-grafana
    restart: always
    ports:
      - "3002:3000"
    networks:
      - bridge-network
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_PASSWORD}
      - GF_USERS_ALLOW_SIGN_UP=false
      - GF_INSTALL_PLUGINS=grafana-clock-panel,grafana-simple-json-datasource
      - GF_SERVER_ROOT_URL=https://metrics.bridge.lux.network
    volumes:
      - grafana-data:/var/lib/grafana
      - ./monitoring/grafana/dashboards:/etc/grafana/provisioning/dashboards
      - ./monitoring/grafana/datasources:/etc/grafana/provisioning/datasources
    deploy:
      resources:
        limits:
          cpus: '1'
          memory: 1G

  # Backup Service
  backup:
    image: ghcr.io/luxfi/bridge-backup:latest
    container_name: bridge-backup
    restart: always
    environment:
      - BACKUP_INTERVAL=3600
      - BACKUP_RETENTION_DAYS=30
      - S3_BUCKET=${BACKUP_S3_BUCKET}
      - AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID}
      - AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY}
    volumes:
      - postgres-data:/data/postgres:ro
      - mpc-data-0:/data/mpc-0:ro
      - mpc-data-1:/data/mpc-1:ro
      - mpc-data-2:/data/mpc-2:ro
    networks:
      - bridge-network

volumes:
  nats-data:
    driver: local
  consul-data:
    driver: local
  mpc-data-0:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /mnt/ssd/bridge/mpc-0
  mpc-data-1:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /mnt/ssd/bridge/mpc-1
  mpc-data-2:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /mnt/ssd/bridge/mpc-2
  postgres-data:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /mnt/ssd/bridge/postgres
  prometheus-data:
    driver: local
  grafana-data:
    driver: local
  bridge-go-data:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /mnt/ssd/bridge/bridge-go

networks:
  bridge-network:
    driver: bridge
    ipam:
      config:
        - subnet: 172.20.0.0/16