You can mass-produce hardware; you cannot mass-produce software - you cannot mass-produce the human mind. –Michio Kaku
Up until now, we used Docker through its client. Whenever we needed something, the only thing we had to do is execute a docker command (example: docker service create). In most cases, that is enough when we are limiting ourselves to operating our cluster from a command line.
What happens if we want to accomplish a functionality beyond what the client offers? What if we'd like to operate Docker from inside our applications? Can we get statistics from all the containers running on the whole cluster?
One possible answer to those and quite a few other questions lies in the adoption of tools beyond those offered by Docker Inc. We'll explore quite a few of those in the chapters that follow.
Another approach would be to use Docker...