Foundations of Software and System Performance Engineering

André B. Bondi is a highly experienced software performance engineer. He founded Software Performance and Scalability Consulting LLC early in 2016. He has recently spent a semester as a visiting professor at the University of L'Aquila in Italy. He spent many years at Siemens Corp., Corporate Technologies in Princeton, New Jersey, and at AT&T Labs and its predecessor, AT&T Bell Labs in Middletown and Holmdel, New Jersey, respectively. He has also held senior performance positions at two startups. In November 2016, Dr. Bondi received the A. A. Michelson Award from the Computer Measurement Group for sustained and valuable contributions to his profession. In addition to holding a doctorate in computer science from Purdue University and an M.Sc. from University College London, Dr. Bondi is a Certified Scrum Master.

Preface xxiii

Acknowledgments xxix

About the Author xxxi

Chapter 1: Why Performance Engineering? Why Performance Engineers? 1

1.1 Overview 1

1.2 The Role of Performance Requirements in Performance Engineering 4

1.3 Examples of Issues Addressed by Performance Engineering Methods 5

1.4 Business and Process Aspects of Performance Engineering 6

1.5 Disciplines and Techniques Used in Performance Engineering 8

1.6 Performance Modeling, Measurement, and Testing 10

1.7 Roles and Activities of a Performance Engineer 11

1.8 Interactions and Dependencies between Performance Engineering and Other Activities 13

1.9 A Road Map through the Book 15

1.10 Summary 17

Chapter 2: Performance Metrics 19

2.1 General 19

2.2 Examples of Performance Metrics 23

2.3 Useful Properties of Performance Metrics 24

2.4 Performance Metrics in Different Domains 26

2.5 Examples of Explicit and Implicit Metrics 32

2.6 Time Scale Granularity of Metrics 32

2.7 Performance Metrics for Systems with Transient, Bounded Loads 33

2.8 Summary 35

2.9 Exercises 35

Chapter 3: Basic Performance Analysis 37

3.1 How Performance Models Inform Us about Systems 37

3.2 Queues in Computer Systems and in Daily Life 38

3.3 Causes of Queueing 39

3.4 Characterizing the Performance of a Queue 42

3.5 Basic Performance Laws: Utilization Law, Little’s Law 45

3.6 A Single-Server Queue 49

3.7 Networks of Queues: Introduction and Elementary Performance Properties 52

3.8 Open and Closed Queueing Network Models 58

3.9 Bottleneck Analysis for Single-Class Closed Queueing Networks 63

3.10 Regularity Conditions for Computationally Tractable Queueing Network Models 68

3.11 Mean Value Analysis of Single-Class Closed Queueing Network Models 69

3.12 Multiple-Class Queueing Networks 71

3.13 Finite Pool Sizes, Lost Calls, and Other Lost Work 75

3.14 Using Models for Performance Prediction 77

3.15 Limitations and Applicability of Simple Queueing Network Models 78

3.16 Linkage between Performance Models, Performance Requirements, and Performance Test Results 79

3.17 Applications of Basic Performance Laws to Capacity Planning and Performance Testing 80

3.18 Summary 80

3.19 Exercises 81

Chapter 4: Workload Identification and Characterization 85

4.1 Workload Identification 85

4.2 Reference Workloads for a System in Different Environments 87

4.3 Time-Varying Behavior 89

4.4 Mapping Application Domains to Computer System Workloads 91

4.5 Numerical Specification of the Workloads 95

4.6 Numerical Illustrations 99

4.7 Summary 103

4.8 Exercises 103

Chapter 5: From Workloads to Business Aspects of Performance Requirements 105

5.1 Overview 105

5.2 Performance Requirements and Product Management 106

5.3 Performance Requirements and the Software Lifecycle 111

5.4 Performance Requirements and the Mitigation of Business Risk 112

5.5 Commercial Considerations and Performance Requirements 114

5.6 Guidelines for Specifying Performance Requirements 116

5.7 Summary 122

5.8 Exercises 123

Chapter 6: Qualitative and Quantitative Types of Performance Requirements 125

6.1 Qualitative Attributes Related to System Performance 126

6.2 The Concept of Sustainable Load 127

6.3 Formulation of Response Time Requirements 128

6.4 Formulation of Throughput Requirements 130

6.5 Derived and Implicit Performance Requirements 131

6.6 Performance Requirements Related to Transaction Failure Rates, Lost Calls, and Lost Packets 134

6.7 Performance Requirements Concerning Peak and Transient Loads 135

6.8 Summary 136

6.9 Exercises 137

Chapter 7: Eliciting, Writing, and Managing Performance Requirements 139

7.1 Elicitation and Gathering of Performance Requirements 140

7.2 Ensuring That Performance Requirements Are Enforceable 143

7.3 Common Patterns and Antipatterns for Performance Requirements 144

7.4 The Need for Mathematically Consistent Requirements: Ensuring That Requirements Conform to Basic Performance Laws 148

7.5 Expressing Performance Requirements in Terms of Parameters with Unknown Values 149

7.6 Avoidance of Circular Dependencies 149

7.7 External Performance Requirements and Their Implications for the Performance Requirements of Subsystems 150

7.8 Structuring Performance Requirements Documents 150

7.9 Layout of a Performance Requirement 153

7.10 Managing Performance Requirements: Responsibilities of the Performance Requirements Owner 155

7.11 Performance Requirements Pitfall: Transition from a Legacy System to a New System 156

7.12 Formulating Performance Requirements to Facilitate Performance Testing 158

7.13 Storage and Reporting of Performance Requirements 160

7.14 Summary 161

Chapter 8: System Measurement Techniques and Instrumentation 163

8.1 General 163

8.2 Distinguishing between Measurement and Testing 167

8.3 Validate, Validate, Validate; Scrutinize, Scrutinize, Scrutinize 168

8.4 Resource Usage Measurements 168

8.5 Utilizations and the Averaging Time Window 175

8.6 Measurement of Multicore or Multiprocessor Systems 177

8.7 Measuring Memory-Related Activity 180

8.8 Measurement in Production versus Measurement for Performance Testing and Scalability 181

8.9 Measuring Systems with One Host and with Multiple Hosts 183

8.10 Measurements from within the Application 186

8.11 Measurements in Middleware 187

8.12 Measurements of Commercial Databases 188

8.13 Response Time Measurements 189

8.14 Code Profiling 190

8.15 Validation of Measurements Using Basic Properties of Performance Metrics 191

8.16 Measurement Procedures and Data Organization 192

8.17 Organization of Performance Data, Data Reduction, and Presentation 195

8.18 Interpreting Measurements in a Virtualized Environment 195

8.19 Summary 196

8.20 Exercises 196

Chapter 9: Performance Testing 199

9.1 Overview of Performance Testing 199

9.2 Special Challenges 202

9.3 Performance Test Planning and Performance Models 203

9.4 A Wrong Way to Evaluate Achievable System Throughput 208

9.5 Provocative Performance Testing 209

9.6 Preparing a Performance Test 210

9.7 Lab Discipline in Performance Testing 217

9.8 Performance Testing Challenges Posed by Systems with Multiple Hosts 218

9.9 Performance Testing Scripts and Checklists 219

9.10 Best Practices for Documenting Test Plans and Test Results 220

9.11 Linking the Performance Test Plan to Performance Requirements 222

9.12 The Role of Performance Tests in Detecting and Debugging Concurrency Issues 223

9.13 Planning Tests for System Stability 225

9.14 Prospective Testing When Requirements Are Unspecified 226

9.15 Structuring the Test Environment to Reflect the Scalability of the Architecture 228

9.16 Data Collection 229

9.17 Data Reduction and Presentation 230

9.18 Interpreting the Test Results 231

9.19 Automating Performance Tests and the Analysis of the Outputs 244

9.20 Summary 246

9.21 Exercises 246

Chapter 10: System Understanding, Model Choice, and Validation 251

10.1 Overview 252

10.2 Phases of a Modeling Study 254

10.3 Example: A Conveyor System 256

10.4 Example: Modeling Asynchronous I/O 260

10.5 Systems with Load-Dependent or Time-Varying Behavior 266

10.6 Summary 268

10.7 Exercises 270

Chapter 11: Scalability and Performance 273

11.1 What Is Scalability? 273

11.2 Scaling Methods 275

11.3 Types of Scalability 277

11.4 Interactions between Types of Scalability 282

11.5 Qualitative Analysis of Load Scalability and Examples 283

11.6 Scalability Limitations in a Development Environment 292

11.7 Improving Load Scalability 293

11.8 Some Mathematical Analyses 295

11.9 Avoiding Scalability Pitfalls 299

11.10 Performance Testing and Scalability 302

11.11 Summary 303

11.12 Exercises 304

Chapter 12: Performance Engineering Pitfalls 307

12.1 Overview 308

12.2 Pitfalls in Priority Scheduling 308

12.3 Transient CPU Saturation Is Not Always a Bad Thing 312

12.4 Diminishing Returns with Multiprocessors or Multiple Cores 314

12.5 Garbage Collection Can Degrade Performance 315

12.6 Virtual Machines: Panacea or Complication? 315

12.7 Measurement Pitfall: Delayed Time Stamping and Monitoring in Real-Time Systems 317

12.8 Pitfalls in Performance Measurement 318

12.9 Eliminating a Bottleneck Could Unmask a New One 319

12.10 Pitfalls in Performance Requirements Engineering 321

12.11 Organizational Pitfalls in Performance Engineering 321

12.12 Summary 322

12.13 Exercises 323

Chapter 13: Agile Processes and Performance Engineering 325

13.1 Overview 325

13.2 Performance Engineering under an Agile Development Process 327

13.3 Agile Methods in the Implementation and Execution of Performance Tests 332

13.4 The Value of Playtime in an Agile Performance Testing Process 334

13.5 Summary 336

13.6 Exercises 336

Chapter 14: Working with Stakeholders to Learn, Influence, and Tell the Performance Engineering Story 339

14.1 Determining What Aspect of Performance Matters to Whom 340

14.2 Where Does the Performance Story Begin? 341

14.3 Identification of Performance Concerns, Drivers, and Stakeholders 344

14.4 Influencing the Performance Story 345

14.5 Reporting on Performance Status to Different Stakeholders 353

14.6 Examples 354

14.7 The Role of a Capacity Management Engineer 355

14.8 Example: Explaining the Role of Measurement Intervals When Interpreting Measurements 356

14.9 Ensuring Ownership of Performance Concerns and Explanations by Diverse Stakeholders 360

14.10 Negotiating Choices for Design Changes and Recommendations for System Improvement among Stakeholders 360

14.11 Summary 362

14.12 Exercises 363

Chapter 15: Where to Learn More 367

15.1 Overview 367

15.2 Conferences and Journals 369

15.3 Texts on Performance Analysis 370

15.4 Queueing Theory 372

15.5 Discrete Event Simulation 372

15.6 Performance Evaluation of Specific Types of Systems 373

15.7 Statistical Methods 374

15.8 Performance Tuning 374

15.9 Summary 375

References 377

Index 385