芹菜工人无法连接到Docker Swarm中的RabbitMQ，但他们在标准Docker compose中工作良好

我试图在单节点VPS上的Docker Swarm中运行我的应用程序。堆栈如下所示：

• 前端：React.js节点使用

  service-s构建提供静态文件
  命令

• 后端：使用gunicorn的Django rest框架API
• RabbitMQ：用作芹菜工人的消息代理
• 芹菜工人：与后端图像相同，但开始芹菜工作 使用

  多启动

  命令
• 芹菜节拍：与后端图像相同，但开始芹菜节拍 排队
• 数据库：使用外部托管的云数据库
• Nginx：使用docker gen的反向代理自动生成 容器的反向代理
• Nginx-letsencrypt：创建SSL证书的代理的伙伴

问题

当我运行

docker-compose-f docker-compose.staging.yml-up--build

时，所有容器都运转良好。芹菜工作者连接到rabbitmq容器并启动工作者，后端连接到数据库，前端提供静态文件，反向代理正确路由请求

但是，当我尝试在群中使用命令

docker stack-c docker-compose.staging.yml deploy app

运行此操作时，芹菜工人不会启动工人。所有芹菜工人带着9号信号离开。芹菜节拍调度程序容器的日志表明，工作容器无法解析代理（RabbitMQ）的主机名

从

docker compose

到

docker swarm

的网络中有哪些变化阻止了工人连接和启动

我的撰写文件（已编辑）-docker compose.staging.yml

version: "3.3"

services:
  # RabbitMQ - queue
  rabbitmq:
    container_name: rabbit
    image: rabbitmq:3-management
    expose:
      - "15672"
      - "5672"
    environment:
      - RABBITMQ_DEFAULT_USER=example
      - RABBITMQ_DEFAULT_PASS=password
      - VIRTUAL_PORT=15672
      - VIRTUAL_HOST=rabbit.example.com
      - LETSENCRYPT_HOST=rabbit.example.com
      - LETSENCRYPT_EMAIL=example@me.com
    restart: on-failure
    deploy:
      restart_policy:
        condition: on-failure
      placement:
        max_replicas_per_node: 1
        constraints:
          - "node.role==manager"

  # Backend API and admin portal
  backend:
    container_name: backend
    image: mycustomprivate/backend-image
    command:
      [
        "gunicorn",
        "--bind",
        "0.0.0.0:8000",
        "main.wsgi:application",
      ]
    entrypoint: /backend/production-entrypoint.sh
    expose:
      - "8000"
    depends_on:
      - rabbitmq
    restart: on-failure
    env_file:
      - ./backend/.env
    environment:
      - DJANGO_SETTINGS_MODULE=main.staging
      - VIRTUAL_PORT=8000
      - VIRTUAL_HOST=api.example.com
      - LETSENCRYPT_HOST=api.example.com
      - LETSENCRYPT_EMAIL=example@me.com
    deploy:
      replicas: 5
      update_config:
        parallelism: 2
      restart_policy:
        condition: on-failure

  # Static files for React.js frontend served
  frontend:
    container_name: frontend
    image: mycustomprivate/frontend-image
    command: serve -s build
    entrypoint: /frontend/staging-entrypoint.sh
    expose:
      - "5000"
    environment:
      - VIRTUAL_HOST=example.com
      - VIRTUAL_PORT=5000
      - LETSENCRYPT_HOST=example.com
      - LETSENCRYPT_EMAIL=example@me.com
    stdin_open: true
    deploy:
      replicas: 5
      update_config:
        parallelism: 2
      restart_policy:
        condition: on-failure

  # # Celery - workers
  celery_worker:
    container_name: celery_worker
    image: mycustomprivate/backend-image
    command:
      [
        "./wait-for-it",
        "rabbitmq:5672",
        "--",
        "../configuration/docker_run_workers.sh",
      ]
    depends_on:
      - backend
      - rabbitmq
    restart: on-failure
    env_file:
      - ./backend/.env
    environment:
      - DJANGO_SETTINGS_MODULE=main.staging
    volumes:
      - ./configuration:/configuration
    deploy:
      replicas: 5
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
        window: 120s
      resources:
        limits:
          cpus: "0.50"
          memory: 50M
        reservations:
          cpus: "0.25"
          memory: 20M

  # # Celery - beat
  celery_beat:
    container_name: celery_beat
    image: mycustomprivate/backend-image
    command:
      [
        "./wait-for-it",
        "rabbitmq:5672",
        "--",
        "../configuration/docker_run_beat.sh",
      ]
    depends_on:
      - backend
      - rabbitmq
      - celery_worker
    restart: on-failure
    environment:
      - DJANGO_SETTINGS_MODULE=main.staging
    volumes:
      - ./configuration:/configuration
    deploy:
      restart_policy:
        condition: on-failure
      placement:
        max_replicas_per_node: 1
        constraints:
          - "node.role==manager"