This post originated from an RSS feed registered with Python Buzz
by Aaron Brady.
Original Post: Nagios: 2nd Edition (No Starch Press)
Feed Title: insom.me.uk
Feed URL: http://feeds2.feedburner.com/insommeuk
Feed Description: Posts related to using Python. Some tricks and tips, observations, hacks, and the Brand New Things.
I’ve just finished reading half of, and thumbing brusquely through the rest of, “Nagios” by Wolfgang Barth.
Inverted pyramids and all that: This book is great. At work I had discovered that, while we were using Nagios, we could probably be doing more with it. I’m a fan of book-learning, when it’s from actual books (sorry trees), so I had a good search around for what books are available.
I’ll leave it to the interested reader to do a search, but basically this book was a little more than the others, but won out based on two things; I have an innate like of O’Reilly books - the editorial quality is generally much higher - and it promised good coverage of setting up a distributed Nagios install.
The introductory and installation chapters, something I’ve written about before are actually useful for a change, explaining how building from source differs from using the respective packages, and what some example compile-time/build-time options are. It lists the basic dependencies for the major distributions (some OpenSuSE bias is present; that seems fair enough).
The half of the book which I actually read concisely explains the various objects you can create in Nagios, how they relate to each other, how the inheritance works and how to build a working set of configuration files.
It explains, and is even apologetic for, the various ways that tests can be run - locally, remotely, over SSH, over NRPE, and waiting for passive NSCA reception - and the pros and cons for each. The author goes to pains not to dismiss any given approach as “wrong” or legacy; though he still gives plenty of useful examples and scenarios.
An obligatory explanation of the included plugins and their options is included, but this isn’t gratuitous page filler because examples of when you would use each option are liberally provided, and the context of when to use the parameters is often missing for Nagios plugins (which do not generally include man pages).
There’s a kind of … middle section on visualisation which I didn’t expect to be as interesting as it was; the idea that for a mid-level installation you could replace Cacti or Munin with Nagios performance data is appealing. Unfortunately as with Nagios in general, TIMTOWTDI, so multiple options are investigated, rather than presenting a preferred one.
The latter half of the book, which I scanned, deals with SNMP, Windows servers, monitoring some specific services (LM Sensors, SAP) and setting up distributed monitoring. These are useful chapters to refer to, but I didn’t think the content mattered enough to me to read the whole thing.
The very final chapters explain how to develop your own plugins, and then the appendices explain the configuration format, how to optimise large installations and the new features in Nagios 3.0.
In conclusion: if you’re planning a new Nagios installation, or you (correctly) suspect yours could be working harder for you, buy this book.
Some amount of the excitement in reading this comes from finding new features I can use, so I guess that’s actually a positive for Nagios, rather than for how the book is written, but then, that’s why I bought the book in the first place.