Ganglia is the most robust and scalable tool for performance monitor I’ve tried or heard of.
This book, written by some of the top contributor of the project, is an awesome guide to Ganglia.
Due to its organization and the authors writing style, the book is easy to understand and can be read as a “full-guide” reading it from the first page to the last one, or a reference book reading only the parts that are relevant to you in that specific moment.
Due to the software nature and history, the authors always speak about scalability and big-number-optimization, leaving the reader that is looking for a simple one-node configuration a little bit confused. The confusion, as far as I can tell, disappears pretty soon. This is the only downside I found in the book: there is a lot of information about big environments but close to none information specific for small or single-node environments.
The two chapters I liked the most are the Chapter 7 (Ganglia and Nagios) and the Chapter 9 (Ganglia Case Studies). The Chapter 7, as the title makes easy to guess, suggests how to make Ganglia and Nagios to pass information one to the other. I believe this is very important since I usually find that both tools are not perfect, but they are (close to be) perfectly complementary. Chapter 9 brings up 6 real-world cases and a lot of information that are really useful if you plan to create an environment somehow similar to one (ore more) of the presented cases.
I would definitely suggest this book to anyone is facing an installation or a configuration of Ganglia or to anyone has to monitor multiple systems in an optimized way.
You can find the book at O’Reilly website.
Disclaimer: I received a free electronic copy of this book as part of the O’Reilly Blogger Program