Distributed Algorithms: An Intuitive Approach Illustrated Edition

I've only glanced through it, but for whatever weight you give that it looks perfect. I say 'for the practitioner', but it's not just a handful of code for the usual algorithms, but rather thorough explanation as well as lots of variations, with plenty of diagrams and pseudo code. Sounds unlikely from a book with just 200 pages, but it's so (see the TOC). It's laid out with academic rigor (no chit-chat), except without proofs. Contains exercises, though no solutions. It's done well enough to look like pleasurable bathroom reading!

How many algorithms your computer is using just to let you read book reviews online? Well, many of them are in this book. Specially those connecting your computer to amazon’s machines. The so called distributed algorithms. Remember that the Internet is the largest distributed system ever. And it is getting even more distributed so this book will help you understand how such a system scales. Fokkink extends the analysis of algorithms to include the bounds of message complexity. One of the goals is compute the cost of using the network reusing the big O notation often used for time and space. And the book goes through several algorithms like a textbook also packed with exercises.

Recommended for anyone interested in algorithms. Author has clearly put a lot of effort into making his explanations understandable. Make no mistake though, it's very dense material. Take it slow and enjoy. Good exercises for checking to see that you understand as much as you think you do. Paper and binding is excellent quality, feels good in my hands. Don't miss the algorithm listing in the back.

Review based on the first 7 chapters.The book skips proofs and just explains why the algorithms work, often with examples and pictures. This I like very much.Caveat: Many of the algorithms given in the book assume a model of distributed computation where messages are not lost and nodes (processes) do not die. In other words, many of these algorithms are not applicable in modern Google-style shared-nothing distributed computing, where you have 1000s of components in a datacenter, and at any one time, a number of them will be failing or unreachable. These are algorithms for multiple processes (processors) in a single computer, where message passing is reliable and processes don't randomly die. That's fine, but I wish the book's title or description would have been clear about this.Since most of my past readings were in algorithms that work in the case of failures and message loss (like Paxos) I did find the book educational.

It’s an amazing outline of algorithms that you should dive further into. Other books provide a more in-depth subset of this information, however, I’d rather get a high level of the entire space and then have the option to dive deeper rather than have a large gap in breadth.

I've read a few sections of this book, but not all of it. From what I could see, this book imposes much less overhead than Lynch or Attiya if you need to understand the basics of an algorithm quickly. It is a good choice for anyone who doesn't want to look through a long prologue of formalism before getting to some of the more useful algorithms.Like other reviewers noted, some of the algorithms do not consider faulty channels and faulty nodes, but this is consistent with the terminology used in books like Lynch or Attiya, where both shared memory systems and shared nothing systems are treated in the same book. From this perspective, faulty channels and nodes are also special case. I think we are still missing a book that focuses and only on the intuition for the vast majority of algorithms used in practice (with the failure modes seen in practice). This would make the book still shorter.

Very disappointed with this book, bought with the high hopes after going through couple of reviews here but the book turned out to be a very boring one. Seemed to contain nothing but a collection of short (published) papers for e.g. Deadlock detection, Termination detection, Leader election and etc. I would have benefited if the authors had spent more time explaining each of the algorithms in detail rather than presenting short description of how they work with minimal or no examples.

The biggest disappointment for me was that exercises in the book come without solutions - not useful for self-study.