Review By: Sunil S. Prasad
06/08/2012

A functional requirement for an e-commerce software system could claim, "The purchase operation must be executable only after the user has been authenticated through the login operation." For the same system, a non-functional requirement could claim, "The average execution time of the purchase operation must be lower than two seconds." But, what is an execution time? What are the parameters that contribute to it? How can it be analyzed?

Model-based Software Performance Analysis talks about techniques, methods, and tools for system performance in software development. Seven beautifully articulated chapters clarify the concept in multiple ways. The book introduces performance concerns into the scope of software models. Developers can carry out performance analysis all along the software lifecycle, and testers can be ready early enough to test happy flow, negative flow, and alternative flow cases. In sum, everyone can be on the same page and there are no surprises at the end of the project.

One of the main causes of poor performance is quality-related shortcomings. The authors have done well to provide examples of e-commerce, use case diagrams, state machine diagrams, and UML profiles that allow developers to fuel performance. It's like the old adage "measure twice before cutting." The developer while designing can translate the software model into performance that will help not only technical teams but also business understand in a simpler way. This is possible by the authors' description of the most representative methodologies used to annotate and transform software models into performance models.

My personal favorite is the last chapter, "Advanced Issues in Software Performance." The authors have done great justice in providing an introduction to this domain and suggesting other books to read to achieve higher level of knowledge. Professional software engineers and graduate students in software engineering can read Model-based Software Performance Analysis to learn both basic concepts of performance modeling and new methodologies.