For Docker, stacks will be the standard way of packaging applications made by multiple containers. Consider the hyper inflated WordPress example: You need a minimum of one web server and one database.

Developers usually describe these applications with Compose, by creating a YAML, as shown:

version: '2'
services:
  db:
    image: mysql:5.6
    volumes:
      - "./.data/db:/var/lib/mysql"
    restart: always
    environment:
      MYSQL_ROOT_PASSWORD: dockerswarm
      MYSQL_DATABASE: wordpress
      MYSQL_USER: wordpress
      MYSQL_PASSWORD: wordpress
  wordpress:
    depends_on:
      - db
    image: wordpress:latest
    links:
      - db
    ports:
      - "8000:80"
    restart: always
    environment:
      WORDPRESS_DB_HOST: db:3306
      WORDPRESS_DB_PASSWORD: wordpress

Then, they launch this application with a command such as:

docker-compose --rm -d --file docker-compose.yml up.

Here, mysql and wordpress containers are scheduled, pulled, and started as daemons...