Andre B. Bondi, the author, has a background in mathematics and statistics, as well as many years of experience in network management. Thus, it is not surprising that his view on system performance is presented from the viewpoint of a quantitative approach and management procedures. While the book’s title emphasizes the software aspect, and it is covered to some extent, the traditional system perspective, that is, the one focused on interactions between various components, is dominant here.

First, the justification for caring about performance engineering is given, along with the background. Then, in chapter 2, metrics used to assess and quantify the performance are discussed. Apart from giving basic definitions and examples, chapter 3 covers queueing modeling, which may be used to predict the values of the metrics for whole systems. Examples are given, so this is not only a presentation of mathematical material. After giving theoretical background, Bondi focuses on the procedural aspects of performance evaluation, starting with a definition of a key notion describing system operation, that is, workload (in chapter 4), and then elaborating on the business approach to system requirements (in chapter 5). Then, in chapter 6, the author focuses on how to define performance requirements in various ways, while in the subsequent chapter he discusses how to find the related values and how to describe and manage system performance. This material is then extended with a chapter on measurement procedures and how to plan and carry out performance tests in various systems. Chapter 10 is devoted to ways of preparing a system model. Chapter 11 focuses on one of the most important factors related to system growth: scalability issues. The next chapter is my favorite, since it presents many issues to avoid while designing systems. Such knowledge is typically most useful for practitioners, and the only negative aspect related to this fragment of the book is that it is not longer. Chapter 13 proposes how to incorporate performance engineering in agile development frameworks. The last chapter discusses how to communicate performance results and assumptions to external groups or bodies of stakeholders who do not necessarily have a technical background. The book finishes with suggestions on where to find additional information on the topic.

The book is aimed at novices in the field of performance evaluation in software, computer, or network systems. While the material presented is not very advanced, Bondi gives many interesting intuitions and examples apparently based on his rich experience; therefore, the book is really a good source of knowledge. Each chapter is appended with exercises or problem questions, hence the work can be used as a textbook.

While there are many books on the performance evaluation of systems, the author’s position presents a good balance between the engineering-focused qualitative books on the topic, and the books overloaded with mathematical modeling that leave readers without clear hints on how to use the information in practice. While there are good books treating these aspects separately, for example [1] dealing with the first aspect or [2] with the other, here, the author combines both perspectives, presenting how to handle issues from a procedural viewpoint, but also giving some quantitative approaches just in time to teach readers how to use even very simple models to gain insight into system operation.