GoClaw/docker/docker-stack.yml

##############################################################################
# GoClaw Control Center — Docker Stack (Docker Swarm Production)
#
# Deploy:
#   docker stack deploy -c docker/docker-stack.yml goclaw
#
# Remove:
#   docker stack rm goclaw
#
# Scale gateway:
#   docker service scale goclaw_gateway=3
##############################################################################

version: "3.9"

networks:
  goclaw-net:
    driver: overlay
    attachable: true

volumes:
  ollama-data:
    driver: local
  mysql-data:
    driver: local

configs:
  gateway-env:
    external: true

secrets:
  mysql-root-password:
    external: true
  mysql-password:
    external: true
  jwt-secret:
    external: true

services:

  # ── MySQL 8 ──────────────────────────────────────────────────────────────
  db:
    image: mysql:8.0
    environment:
      MYSQL_ROOT_PASSWORD_FILE: /run/secrets/mysql-root-password
      MYSQL_DATABASE: goclaw
      MYSQL_USER: goclaw
      MYSQL_PASSWORD_FILE: /run/secrets/mysql-password
    secrets:
      - mysql-root-password
      - mysql-password
    volumes:
      - mysql-data:/var/lib/mysql
    networks:
      - goclaw-net
    deploy:
      replicas: 1
      placement:
        constraints:
          - node.role == manager
      restart_policy:
        condition: on-failure
        delay: 10s
      resources:
        limits:
          memory: 1G
        reservations:
          memory: 512M
    healthcheck:
      test: ["CMD", "mysqladmin", "ping", "-h", "localhost"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 30s

  # ── Ollama LLM Server ─────────────────────────────────────────────────────
  ollama:
    image: ollama/ollama:latest
    volumes:
      - ollama-data:/root/.ollama
    networks:
      - goclaw-net
    deploy:
      replicas: 1
      placement:
        constraints:
          # Pin to GPU node if available
          - node.labels.gpu == true
      restart_policy:
        condition: on-failure
        delay: 15s
      resources:
        limits:
          memory: 16G
        reservations:
          memory: 4G
    # GPU support via nvidia-container-runtime
    # Uncomment on GPU-enabled nodes:
    # runtime: nvidia
    # environment:
    #   - NVIDIA_VISIBLE_DEVICES=all

  # ── Go Gateway (Orchestrator + Tool Executor) ─────────────────────────────
  gateway:
    image: git.softuniq.eu/uniqai/goclaw/gateway:latest
    environment:
      PORT: "18789"
      OLLAMA_BASE_URL: "http://ollama:11434"
      DATABASE_URL: "goclaw:{{MYSQL_PASSWORD}}@tcp(db:3306)/goclaw?parseTime=true"
      PROJECT_ROOT: "/app"
      REQUEST_TIMEOUT_SECS: "120"
      LOG_LEVEL: "info"
    secrets:
      - mysql-password
    networks:
      - goclaw-net
    ports:
      - target: 18789
        published: 18789
        protocol: tcp
        mode: ingress
    deploy:
      replicas: 2
      update_config:
        parallelism: 1
        delay: 10s
        order: start-first
        failure_action: rollback
      rollback_config:
        parallelism: 1
        delay: 5s
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
      resources:
        limits:
          memory: 512M
          cpus: "1.0"
        reservations:
          memory: 128M
          cpus: "0.25"
    healthcheck:
      test: ["CMD", "wget", "-qO-", "http://localhost:18789/health"]
      interval: 15s
      timeout: 5s
      retries: 3
      start_period: 10s

  # ── Control Center (React + Node.js) ─────────────────────────────────────
  control-center:
    image: git.softuniq.eu/uniqai/goclaw/control-center:latest
    environment:
      NODE_ENV: production
      DATABASE_URL: "mysql://goclaw:{{MYSQL_PASSWORD}}@db:3306/goclaw"
      GATEWAY_URL: "http://gateway:18789"
    secrets:
      - mysql-password
      - jwt-secret
    networks:
      - goclaw-net
    ports:
      - target: 3000
        published: 3000
        protocol: tcp
        mode: ingress
    deploy:
      replicas: 2
      update_config:
        parallelism: 1
        delay: 10s
        order: start-first
        failure_action: rollback
      rollback_config:
        parallelism: 1
        delay: 5s
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
      resources:
        limits:
          memory: 1G
          cpus: "1.0"
        reservations:
          memory: 256M
          cpus: "0.25"
    healthcheck:
      test: ["CMD", "wget", "-qO-", "http://localhost:3000/api/health"]
      interval: 15s
      timeout: 5s
      retries: 3
      start_period: 20s

  # ── Traefik Reverse Proxy (optional) ─────────────────────────────────────
  # traefik:
  #   image: traefik:v3.0
  #   command:
  #     - "--providers.docker.swarmMode=true"
  #     - "--providers.docker.exposedbydefault=false"
  #     - "--entrypoints.web.address=:80"
  #     - "--entrypoints.websecure.address=:443"
  #     - "--certificatesresolvers.letsencrypt.acme.email=admin@softuniq.eu"
  #     - "--certificatesresolvers.letsencrypt.acme.storage=/letsencrypt/acme.json"
  #     - "--certificatesresolvers.letsencrypt.acme.httpchallenge.entrypoint=web"
  #   ports:
  #     - "80:80"
  #     - "443:443"
  #   volumes:
  #     - /var/run/docker.sock:/var/run/docker.sock:ro
  #     - traefik-certs:/letsencrypt
  #   networks:
  #     - goclaw-net
  #   deploy:
  #     placement:
  #       constraints:
  #         - node.role == manager