Preface
Goal, Theme, and Approach
The Internet is profoundly affecting almost all businesses and commerce paradigms. Global competition, industry traditions, laws, and consumer preferences are among the many issues being impacted by e-business. Weekly magazines and newspapers have dedicated large amounts of space to discuss e-business and the new economy. Newspapers publish news about e-business almost every day. However, there is good news and bad news. The good news is many innovative people are constantly creating new and exciting ways to use the Internet to provide novel, efficient, and convenient shopping paradigms as well as streamlining ways for businesses to do business with their partners. The bad news is the most successful e-business sites are at greater risks of being overwhelmed by large numbers of customers who can potentially bring the site operation to a halt. Scalability is then a challenge to the success of e-business. Our primary goal in writing this book was to create a framework in which e-business scalability could be discussed and evaluated.
This book presents a quantitative approach to understanding and analyzing e-business scalability based on a four-level reference model. The four-level model is composed of a business model, a functional model, a customer behavior model, and an IT resource model. This framework is used throughout the book to explain how e-business technologies work and how they impact performance, to characterize and forecast the workload of e-business sites, and to plan their capacity with the use of performance models.
New performance metrics for e-business are presented in the book. Models at the various levels of the reference model are used to represent and understand problems in e-business. These models include customer behavior models (e.g., Customer Behavior Model Graphs and Customer Visit Models), Client/Server Interaction Diagrams, and analytic (e.g., state transition diagrams and queuing networks) and simulation performance models. The combined use of these models provides a framework for assessing and evaluating the scalability of e-business sites.
Many examples derived from real e-business situations are used to illustrate the concepts presented in the book.
Who Should Read this Book
This book can be used by graduate and senior-level computer science students as well as students in e-commerce programs offered by business schools. Students of MBA programs with a concentration in information technology will also benefit from the concepts presented in this book.
Many professionals can use this book as a reference or as a way to learn about e-commerce technologies and the quantitative methods used to evaluate and size e-commerce sites. Examples of such professionals include Web-masters of e-commerce sites and ISPs, CIOs, system architects, project managers of e-commerce companies, e-commerce application developers, capacity planners and performance analysts, designers of Internet-based products for e-commerce, and e-commerce consultants.
Book Organization
Part I: Modeling for E-Business
Chapter 1 presents a framework for analyzing and designing e-business sites. The framework is based on a four-level model that includes a business model, a functional model, a customer behavior model, and an IT resource model. The chapter discusses various types of electronic markets including various types of business-to-business, business-to-consumer, and consumer-to-consumer markets.
Chapter 2 discusses the motivations for capturing user behavior and presents customer models at various levels. The models presented in this chapter can be used to answer what-if questions at the customer behavior level and can serve as a basis to build models at the resource level. Two types of customer models are presented in this chapter: the Customer Behavior Model Graphs and the Customer Visit Model.
Chapter 3 discusses the anatomy of e-business functions from the logical view point. The chapter describes how software servers interact to implement e-business functions and provides a graphical notation—the Client/Server Interaction Diagrams—to describe these interactions. Several numerical examples illustrate how these diagrams can be used to answer quantitative questions about the execution of e-business functions.
Part II: Evaluating E-Business Infrastructure and Services
Chapter 4 describes the various elements that comprise the IT infrastructure needed to support e-businesses. The discussion includes hardware, software, and networking issues. Multi-tier website architectures, composed of dynamic load balancers, Web servers, application servers, and database servers, are described.
Chapter 5 presents a quantitative analysis of authentication services used in e-business. The chapter shows how the mechanisms and protocols used to support security may impact system performance. Many quantitative examples illustrate the tradeoffs between performance and security. In particular, the chapter analyzes how authentication protocols such as the Transport Layer Security (TLS) protocol, a successor of the Secure Sockets Layer (SSL) protocol, affects performance.
Chapter 6 provides an overview of what happens when one uses a credit card for payment in the physical world. Then, it discusses how Secure Electronic Transaction (SET) allows for credit card payments to take place over the Internet. The chapter describes SET at a high enough level of detail to provide the reader with an overall picture of the protocol as well as its performance implications. The performance of SET transactions is then discussed through various numerical examples. The chapter concludes with a brief discussion of other payment services.
Part III: Capacity Planning for E-Business
Chapter 7 contains an example-driven description of a capacity planning methodology for e-businesses. The methodology is composed of three planning activities: business and functional planning, customer behavior planning, and IT resource planning. Each of these activities is described in detail and illustrated with examples.
Chapter 8 provides insight and intuition regarding how simple performance models can be constructed, solved, and used in the context of electronic business environments. The fundamentals of performance models are then introduced. These include concepts such as service time, service demand, waiting time, response time, throughput, and performance laws. Scalability analysis techniques based on the study of performance bounds is presented here.
Chapter 9 starts by introducing very simple models. Complexity is progressively introduced and the solution to each model is presented using first principles and intuitive concepts. Two broad categories of models are covered in this chapter: system- and component-level models. System-level models treat the actual system as a black box that receives requests, processes them, and returns the results. Component-level models, based on queuing networks, in which components are represented by queues, allows one to explicitly represent processors, storage subsystems, networks, and routers.
Chapter 10 uses several examples to illustrate the performance impacts of contention for software resources (e.g., threads of a software server, database locks, and semaphores). Performance models that deal with software contention may be based on approximate analytic models or on simulation models. This chapter gives an overview of the techniques used to solve approximations that represent the effects of software contention. To show the impact of software contention, we discuss various results obtained by a combination of simulation and analytic models.
Chapter 11 presents a methodology for characterizing e-business workloads. In particular, the chapter shows how Customer Behavior Model Graphs (CBMGs) and Customer Visit Models (CVMs) can be obtained from HTTP logs and describes methods, based on clustering analysis, to derive small groups of CBMGs or CVMs that accurately represent the workload. The chapter also shows how the parameters for the resource models (e.g., queuing network models) can be derived from customer behavior models.
Chapter 12 discusses techniques that can be used to analyze and forecast the demand for e-business sites. The chapter discusses traffic burstiness, traffic patterns in e-business, and forecasting techniques including regression methods, moving averages, and exponential smoothing. Logs of a real e-tailer are used to illustrate the traffic characteristics of an e-business site.
Part IV: Models for Specific E-Business Segments
Chapter 13 illustrates the use of the quantitative methods presented in the book through several examples in the business-to-consumer segment. A hypothetical electronic retailer is described. Then, different planning situations are discussed. In light of the models introduced throughout the book, we show how to tackle the problems and we present their solutions. The goal of the chapter is to guide the reader, in a step-by-step manner, through the model-based solution of a number of e-commerce examples.
Chapter 14 illustrates the use of the quantitative methods presented in the book through several examples in the business-to-business segment. An example of supply-chain integration and made-to-order computers is used to illustrate how capacity planning can be carried out in business-to-business environments. The chapter also shows how business-to-consumer transactions generate business-to-business transactions.
Part V: Summary, Challenges, and Perspectives
Chapter 15 concludes the book with a summary of important points covered in the book, challenges for e-business, and future perspectives.
Appendix A contains a glossary of the important terms introduced in the book.
Acknowledgments
The authors would like to thank their many colleagues for discussions that contributed substantially to this book. Special thanks go to our co-author in the first book of this series, Larry Dowdy of Vanderbilt University, for his enthusiasm and dedication to the field of performance evaluation. Thanks go to Rodrigo Fonseca, Wagner Meira Jr., Marco Aurelio Mendes, Goedson Paixão, Adriano Pereira, and Flavia Peligrinelli Ribeiro, who collaborated on several papers on e-commerce with the authors of this book, to Ross Lumley who pointed to information on SET, and to Krysztof Gaj for providing useful literature on cryptography. Particular thanks go to Stuart Feldman of the IBM Institute for Advanced Commerce, Howard Frank of the Robert H. Smith School of Business at the University of Maryland at College Park, and to Jim Gray of Microsoft Research for their praise of our work and permission to quote them in the book. The authors would like to express their most sincere appreciation to Jim Gray for writing the foreword for the book and for providing useful references to papers and reports.
Working with the team at Prentice Hall PTR was a real pleasure. We would like to thank our editor, Tim Moore, for his enthusiasm and continued support throughout this project, for being always available to discuss issues and problems, and for his ideas regarding the packaging of the book. The competence of our production manager, Joan L. McNamara, was essential in making it possible to produce this book in such a short time. We would like to thank Lori Hughes for her superb job as copyeditor and for designing the LaTeX macros we used to prepare the book. Thanks also go to the art department for their work on the cover of the book and to many others at PTR who made this project possible.
Daniel Menascé would like to thank his students and colleagues at George Mason University for providing a stimulating work environment. He would also like to thank the Office of the Provost and the Chair of the Computer Science Department, Henry Hamburger, at George Mason University for awarding him a study leave during the Fall of 1999, which allowed him to dedicate more time to this book. Daniel would like to recognize his mother and the much cherished memory of his father for all the love and direction in life he received from them. Special thanks go to his wife Gilda for a life full of love, dedication, companionship, encouragement, and support, and for having given him two real treasures, his children Flavio and Juliana, whose love, friendship, and positive attitude in life have been a continuous delight.
Virgilio Almeida would like to thank his colleagues, students, and staff at the Computer Science Department of the Federal University of Minas Gerais (UFMG). He would also like to thank CNPq (the Brazilian Council for Scientific Research and Development) and the SIAM/Finep Project, which provided partial support for his work on this book. Virgilio would also like to thank his wife Rejane and sons Pedro and Andre for their joy, patience, and encouragement during the many nights and weekends he worked on this book.
From the Same Authors
Menascé and Almeida co-authored the following books:
Capacity Planning for Web Performance: Metrics, Models, and Methods, D. A. Menascé and V. A. F. Almeida, Prentice Hall, 1998.
Capacity Planning and Performance Modeling: From Mainframes to Client-Server Systems, D. A. Menascé, V. A. F. Almeida, and L. W. Dowdy, Prentice Hall, 1994.
Book's Website and Authors' Addresses
Readers of this book can download the various MS Excel workbooks referenced in the various chapters from http://www.cs.gmu.edu/~menasce/ebook/. This website will also be used to keep the readers informed about new developments related to the book. The authors' e-mail and postal addresses and websites are:
Prof. Daniel A. Menascé
Department of Computer Science, MS 4A5
George Mason University
Fairfax, VA 22030-4444
+1 703 993-1537
[email protected]
http://www.cs.gmu.edu/faculty/menasce.html
Prof. Virgilio A. F. Almeida
Department of Computer Science
Universidade Federal de Minas Gerais
P.O. Box 920
31270-010 Belo Horizonte, MG
Brazil
+55 31 499-5887
[email protected]
http://www.dcc.ufmg.br/~virgilio