Using Docker

This is just about the most you could expect from a book about Docker. It's far, far better than the documentation, it talks about best principles and practices, and every single suggestion is accompanied by detailed commands to type to accomplish the tasks. The book has a great throughline of developing and improving a small application, which is used as a reference in every single chapter.

It covers almost everything you want to know about Docker including more operational stuff like monitoring, orchestration, and security. Pretty much no stone is left unturned, what you want to know about using Docker for real is covered eventually.

Occasionally I wished the book went into a bit more depth, or I felt like the author was sort of hand-wavey. There were a lot of times when he'd say something like "don't do this in production" without much explanation of what you WOULD want to do in production. Additionally, there were frequent mentions of how this or that aspect of Docker will be fixed or changed soon, leaving the reader to both wonder if Docker is truly ready for prime time, as well as if this was the proper time to publish a book on the subject with the technology so in-flux.

My biggest complaint is actually about Docker itself. By later chapters, Mouat would show incredibly complex and lengthy commands to type that left me wondering if it was even possible for documentation to be good enough to get a person to the point where they'd be able to type that command from scratch. What would you do if you wanted to do something slightly different than Mouat's example? It seemed almost impenetrable at times.

One thing I felt was lacking was an example of persistent data storage. The identidock example that runs through the entire book is all non-persistent, every single container could be restarted with complete loss of state and have no effect. The closest thing to a database used, redis, was used as a simple cache that could be emptied upon restart with no penalty. But many applications need real-live databases that need to persist across container restarts, and there's basically no coverage of how one would accomplish such a thing. From my own Googling it looks like the agreed-upon pattern is a data-only container volume linked to the system needing to store data, but it would have been nice if the example application could have incorporated such a fundamentally basic need.

Overall the book is fair and does a great job of helping the reader understand the pros and cons of containerized deployments, and it takes you through pretty much everything to know, at least in the world of Docker. In terms of how you might want to change your applications to make them more containerizable, it's largely left to the reader. Nonetheless, I can't imagine there even being a book on Docker that's more thorough and properly-paced than this one, it's basically a Missing Manual.