Lean Computing for the Cloud

Applies lean manufacturing principles across the cloud service delivery chain to enable application and infrastructure service providers to sustainably achieve the shortest lead time, best quality, and value Applies lean thinking across the cloud service delivery chain to recognize and minimize waste Leverages lessons learned from electric power industry operations to operations of cloud infrastructure Applies insights from just-in-time inventory management to operation of cloud based applications Explains how traditional, Information Technology Infrastructure Library (ITIL) and Enhanced Telecom Operation Map (eTOM) capacity management evolves to lean computing for the cloud  

Table of Contents:
Introduction xi Acknowledgments xv Abbreviations xvii 1. Basics 1 1.1 Cloud Computing Fundamentals 1 1.2 Roles in Cloud Computing 6 1.3 Applications 9 1.3.1 Application Service Quality 11 1.4 Demand, Supply, Capacity, and Fungibility 13 1.5 Demand Variability 16 1.6 Chapter Review 18 2. Rethinking Capacity Management 19 2.1 Capacity Management 19 2.2 Demand Management 21 2.3 Performance Management 21 2.4 Canonical Capacity Management 23 2.4.1 Traditional Capacity Management 24 2.4.2 ITIL Capacity Management 27 2.4.3 eTOM Capacity Management 28 2.4.4 Discussion 30 2.5 Three Cloud Capacity Management Problems 30 2.5.1 Physical Resource Capacity Management 31 2.5.2 Virtual Resource Capacity Management 32 2.5.3 Application Capacity Management 33 2.6 Cloud Capacity Management as a Value Chain 36 2.7 Chapter Review 39 3. Lean Thinking on Cloud Capacity Management 41 3.1 Lean Thinking Overview 41 3.2 Goal 42 3.3 Seeing Waste (Nonvalue-Adding Activities) 43 3.3.1 Reserve Capacity 45 3.3.2 Excess Application Capacity 46 3.3.3 Excess Online Infrastructure Capacity 46 3.3.4 Excess Physical Infrastructure Capacity 46 3.3.5 Inadequate Capacity 47 3.3.6 Infrastructure Overhead 48 3.3.7 Capacity Management Overhead 48 3.3.8 Resource Overhead 49 3.3.9 Power Management Overhead 50 3.3.10 Workload Migration 50 3.3.11 Complexity Overhead 51 3.3.12 Resource Allocation Failure 51 3.3.13 Leaking and Lost Resources 53 3.3.14 Waste Heat 53 3.3.15 Carbon Footprint 54 3.4 Key Principles 54 3.4.1 Move toward Flow 55 3.4.2 Pull versus Push 55 3.4.3 Level the Workload 55 3.4.4 Stop and Fix Problems 55 3.4.5 Master Practices 56 3.4.6 Visual Management 57 3.4.7 Use Well-Tested Technology 57 3.4.8 Take a Long-Term Perspective 58 3.4.9 Grow, Learn, and Teach Others 58 3.4.10 Develop Exceptional People 58 3.4.11 Partners Help Each Other Improve 58 3.4.12 Go See 59 3.4.13 Implement Rapidly 59 3.4.14 Become a Learning Organization 59 3.5 Pillar: Respect 59 3.6 Pillar: Continuous Improvement 61 3.7 Foundation 62 3.8 Cadence 62 3.9 Lean Capacity Management Philosophy 63 3.10 Chapter Review 64 4. Lean Cloud Capacity Management Strategy 67 4.1 Lean Application Service Provider Strategy 68 4.1.1 User Workload Placement 71 4.1.2 Application Performance Management 73 4.2 Lean Infrastructure Service Provider Strategies 73 4.2.1 Physical Resource Capacity Management 76 4.3 Full Stream Optimization 77 4.4 Chapter Review 79 5. Electric Power Generation as Cloud Infrastructure Analog 81 5.1 Power Generation as a Cloud Infrastructure Analog 81 5.2 Business Context 83 5.3 Business Structure 86 5.4 Technical Similarities 88 5.5 Impedance and Fungibility 91 5.6 Capacity Ratings 94 5.7 Bottled Capacity 95 5.8 Location of Production Considerations 95 5.9 Demand Management 97 5.10 Demand and Reserves 98 5.11 Service Curtailment 99 5.12 Balance and Grid Operations 100 5.13 Chapter Review 103 6. Application Capacity Management as an Inventory Management Problem 105 6.1 The Application Capacity Management Service Delivery Chain 105 6.2 Traditional Application Service Production Chain 107 6.3 Elasticity and Demand-Driven Capacity Management 108 6.4 Application Service as Retail Analog 110 6.4.1 Locational Consideration 112 6.4.2 Inventory and Capacity 112 6.4.3 Service Level 113 6.4.4 Inventory Carrying Costs 114 6.4.5 Inventory Decision, Planning, and Ordering 115 6.4.6 Agility 118 6.4.7 Changing Consumption Patterns 118 6.5 Chapter Review 118 7. Lean Demand Management 119 7.1 Infrastructure Demand Management Techniques 120 7.1.1 Resource Scheduling 121 7.1.2 Resource Curtailment 121 7.1.3 Mandatory Demand Shaping 122 7.1.4 Voluntary Demand Shaping 123 7.1.5 Scheduling Maintenance Actions 123 7.1.6 Resource Pricing 123 7.2 Application Demand Management Techniques 124 7.2.1 Queues and Buffers 124 7.2.2 Load Balancers 124 7.2.3 Overload Controls 125 7.2.4 Explicit Demand Management Actions 125 7.2.5 Scheduling Maintenance Actions 125 7.2.6 User Pricing Strategies 126 7.3 Full Stream Analysis Methodology 126 7.3.1 Analyze Applications' Natural Demand Patterns 127 7.3.2 Analyze Applications' Tolerances 128 7.3.3 Create Attractive Infrastructure Pricing Models 129 7.3.4 Deploy Optimal Infrastructure Demand Management Models 130 7.4 Chapter Review 131 8. Lean Reserves 133 8.1 What Is Reserve Capacity? 133 8.2 Uses of Reserve Capacity 135 8.2.1 Random Demand Peaks 135 8.2.2 Component or Resource Failure 136 8.2.3 Infrastructure Element Failure 136 8.2.4 Infrastructure Resource Curtailment or Demand Management Action 137 8.2.5 Demand Exceeding Forecast 137 8.2.6 Lead Time Demand 137 8.2.7 Catastrophic Failures and Force Majeure Events 139 8.3 Reserve Capacity as a Feature 139 8.4 Types of Reserve Capacity 140 8.4.1 Automatic Infrastructure Power Management Controls 140 8.4.2 Utilize Application Reserve Capacity 141 8.4.3 Place/Migrate Demand into Underutilized Capacity 141 8.4.4 Grow Online Capacity 141 8.4.5 Service Curtailment/Degradation 141 8.4.6 Mandatory Demand Shaping 141 8.4.7 Voluntary Demand Shaping 142 8.4.8 Emergency Reserves 142 8.5 Limits of Reserve Capacity 144 8.6 Ideal Reserve 144 8.6.1 Normal (Co-located) Reserve 144 8.6.2 Emergency (Geographically Distributed) Reserve 146 8.7 Chapter Review 147 9. Lean Infrastructure Commitment 149 9.1 Unit Commitment and Infrastructure Commitment 150 9.2 Framing the Unit Commitment Problem 151 9.3 Framing the Infrastructure Commitment Problem 153 9.4 Understanding Element Startup Time 155 9.5 Understanding Element Shutdown Time 157 9.6 Pulling It All Together 160 9.7 Chapter Review 166 10. Lean Cloud Capacity Management Performance Indicators 167 10.1 Perfect Capacity Metrics 168 10.2 Capacity Management Metrics 172 10.3 Infrastructure Commitment Metrics 173 10.4 Waste Metrics 174 10.4.1 Reserve Capacity Waste Metrics 174 10.4.2 Excess Application Capacity Metrics 175 10.4.3 Excess Online Infrastructure Capacity Metrics 175 10.4.4 Excess Physical Infrastructure Capacity Metrics 175 10.4.5 Inadequate Capacity Metrics 175 10.4.6 Infrastructure Overhead Waste Metrics 176 10.4.7 Capacity Management Overhead Waste Metrics 176 10.4.8 Resource Overhead Waste Metrics 176 10.4.9 Power Management Overhead Waste Metrics 177 10.4.10 Workload Migration Metrics 177 10.4.11 Complexity Overhead Metrics 178 10.4.12 Resource Allocation Failure Metrics 178 10.4.13 Leaking and Lost Resources 179 10.4.14 Waste Heat Metrics 179 10.4.15 Carbon Footprint Metrics 180 10.5 Key Principle Indicators 180 10.6 Cost of Poor Quality 181 10.7 Metrics and Service Boundaries 182 10.8 Measurements and Maturity 183 10.9 Chapter Review 185 11. Summary 187 11.1 Cloud Computing as a Service Delivery Chain 187 11.2 Lean Cloud Computing 190 11.3 Reimagining Cloud Capacity 192 11.4 Lean Demand Management 195 11.5 Lean Reserves 197 11.6 Lean Infrastructure Service Provider Considerations 198 11.7 Lean Application Service Provider Considerations 198 11.8 Lean Infrastructure Commitment 199 11.9 Visualizing Perfect Capacity 201 11.10 Lean Cloud Computing Metrics 203 11.11 Concluding Remarks 204 References 207 About the Author 211 Index 213