Finally, a couple of words on the history of this book and a note on how astonishingly fast is the development of Docker.

When the project of writing a book on Docker Swarm was just drafted, at the day there was only the old Docker Swarm standalone mode, where a Swarm container was responsible for orchestrating infrastructures of containers, having to rely on external discovery systems, such as Etcd, Consul, or Zookeeper.

Looking back at these times, just some months ago, is like thinking to prehistory. Just later in June, when SwarmKit was open sourced as an orchestration kit and it was included into the Engine as Swarm Mode, a major step ahead was made by the Docker in terms of orchestration. A full, scalable and secure by default, and easy way to orchestrate Docker natively was released. Then, it turned out that the best way of orchestrating Docker was just Docker itself.

But when Infrakit was open sourced in October 2016, a new big step ahead was done in terms of infrastructure...