Customer Reviews

The Art of Application Performance Testing: Help for Programmers and Quality Assurance

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

 

 

This well-written primer provides just enough information to help one get started with automated performance testing. It outlines several recommended practices to automating performance testing (e.g., user activity recording, deploying software monitors, etc) and describes how some of them were put to use in two case studies and how they helped in root cause analyses. It discusses desirable capabilities to look for in performance testing tools and only briefly mentions special considerations for technologically complex applications such as those that use AJAX, SAP, or Citrix, for example. A more in-depth discussion of those considerations and how to address them would have been really useful. Although information on select tool vendors and the type of tool offerings they provide is included, there is no additional description of any of the tools' capabilities or qualities (for this, the reader has to visit the provided vendor's website). Finally, most of the sample scenarios discussed in the book pertain to performance testing in Windows environments; while this won't matter to most readers because the principles discussed are equally applicable to other OS environments, this may matter to some readers who do not work with Windows environments at all.

Though this book is geared towards Quality Assurance testers, developers can take away a lot from this material. The book explains methodologies at a high level and remains vendor-agnostic which is nice, but the context always seems to hover around Windows. If it were not for the specific case studies (which seem like they are from the author's personal experience) I would have to say this book would be more on the boring side.

The author did a great job of "teaching a man to fish" instead of "giving a man a fish". For example, in chapter 2, instead of simply giving a list of tools and telling readers to try them, the author goes into details about what to look for in a performance testing tool and how to choose one based on the scenario.

I would recommend this text to anyone who is new to testing as a great introduction, but I would suspect that experienced testers would also gain some insight from it as well.

I am building a new performance testing practice at my company. This book is the most comprehensive writing I have found on the topic of application performance testing. The book flows well and is replete with valuable metrics and graphs that have been of great benefit to the task at hand. Additionally, the book will serve as a beneficial source of reference for future performance testing engagements.

Well done.

I recommend this book to any Quality Assurance specialist who is involved in performance testing. It provides details on how to effectively plan and conduct performance testing. While the book helps to define a test strategy, it also explain how to analyze the performance test results.

As I've mentioned in my reviews of Effective Software Testing: 50 Specific Ways to Improve Your Testing and Implementing Automated Software Testing: How to Save Time and Lower Costs While Raising Quality, I recently became the manager of my company's automated and performance testing team. My background had previously been in application development, and I purchased this book to get an overview to performance testing. For the most part The Art of Application Performance Testing was a good choice. The core of the book is in chapters 2-4 where the author discusses the fundamentals of performance testing, the process of performance testing, and analyzing the results of a performance test. There were times when I didn't exactly follow some Molyneaux's thought process so I didn't get 100% of the material. This meat and potatoes section of the book is also only about 90 pages so it wouldn't have hurt to have a little more explanation. The biggest example of this was the example he gave where he calculated the virtual user numbers for an application that was going to have an increase in concurrent users. The table with the current production numbers and the numbers for the test seemed to drop out of the air with minimal explanation. This surprised me a bit because I rarely have difficulty following technical books. The other issue I had with this book is one that should never happen. A number of the tables and charts used fonts that were too small to be readable. I have 20/20 near vision, and I have never had a problem with a technical book like this. The final consideration I would offer readers is the price in comparison to the size of the book. The chapter material tops out at 114 pages, and I would say that the cover price of the book is probably overpriced by about 30-40%. Perhaps the author thought that the lack of already published material on this subject merited the higher cost.

It may seem that my comments to mesh with the number of starts that I gave the book. In the end the two material issues I had with the book did not keep me from achieving the purpose I set out with when I bought the book. I know have a good understanding of performance testing so it was worth the investment even if I consider it to have been a little higher than necessary. People who do performance testing as their day to day activity will only find a base level of knowledge here. This is not a stopping point if you are looking to build up your performance testing skills. I would almost put this in the "for managers" category.

Overall: B-