Customer Reviews

Pro JMX: Java Management Extensions (Expert's Voice)

5 star:		(0)
4 star:		(2)
3 star:		(1)
2 star:		(0)
1 star:		(2)

 

 

I wasn't completely new to JMX when I picked up this book. Yet, I felt uncomfortable with how fast the author dove into details that would be relevant to someone who's about to write a JMX implementation, not to a regular user of the technology.

The book's structure is also a bit inconsistent, in my opinion. For example, the author jumps into describing the history and patterns of distributed management system design in chapter 5 -- after two introductory chapters and two very code-focused chapters.

As an advanced manual to itty bitty details of how connectors etc. are implemented, this book is excellent. It's too bad that the back cover doesn't give any indication of this focus. If you're looking for a guide to learn or "just" use JMX, I'd suggest looking elsewhere.

This book does not offer many original or insightful ideas about JMX. In fact, a good portion of the book is copied and pasted from the JSR-160 specification. I recommend downloading that and reading it if you're interested in JMX Remoting. The best book on JMX (although a bit dated) is still by Fleury & Lindfors.

The main problem with this books is its lack of concepts.
It tries to delve direcly into code. But then the code is also very poorly written. None of the code is written to explain the concept.
It feels more like the author cut pasted it from somewhere else.
Some of the concepts - like relationships and roles - leave one wondering with the question, whether the author himself has understood them properly.

A few years ago, JMX emerged as a bright new idea in the Java world. It was crafted as a 3 level structure, as Hanson explains. The lowest being the instrumentation, populated with MBeans that map to resources of some kind, hardware or software. Above these is the agent level and then distributed services.

The book goes into lots of detail about how to fill these levels. So that distributed systems might be made that control resources scattered across the Internet. All this has given rise to startups, often of a very specialised nature, like Intersperse, that make such applications.

However, it's been over a year since this book came out. A countervailing trend seems to have emerged. As Web Services have been studied, the Web Services Description Language was made and its limitations discovered. So Business Process Execution Language has recently arisen to describe business logic, controlling loosely coupled distributed Web Services. So there is competition between JMX and BPEL/WSDL, with the latter nudging ahead. Yes, there are differences in scope. But they overlap enough in the crucial idea of distributed business logic.

So consider whether you want to go ahead with this book or try BPEL/WSDL.

The author looks at JMX from the perspective of IT operations management as well as software developerment. The chapter covering the history of management specifications and technologies makes this book worth the price just on those merits alone.