Capacity Planning for Computer Systems

AcknowledgmentsPreface1 Capacity Planning Foundations What Is Capacity Planning? Capacity Planning—Art and Science The Planning Process Evaluation of Future Scenarios—The Spectrum of Probabilities Capacity Planning and Predicting the Future Technical Factors Business Factors Management Factors Capacity Planning and Performance Management Capacity Planning and Business Planning References2 The Metrics of Capacity Planning—Foundations Fundamentals Resource Categorization Resource Consumption Metrics Resource Usage and Performance Level Metrics for Capacity Projections Relative I/O Content (R) of Workloads Access Density (AD) of Disk Resident Data Process Density of DASD Resident Data (PDD) Process Density of Memory Resident Data (PDM) Data Remoteness Factor (DRF) Memory Access Density (MAD) Resource Exhaustion References3 Business Elements and Service Levels Methodology Identifying Business Elements Statistical Analysis and Modeling Validating the Business Element Statistical Model Service Levels and Objectives Examples of Service Goals Service-Level Agreements Award Points Cost Controls and Chargeback References4 Forecasting, Predictions, and Planning Planning vs. Forecasting Statistical Forecasting Probability and the Analysis of Risk Contingency—System Failure and Recovery Prediction, Forecasting, and Visualization References5 Statistical Forecasting for Capacity Planning Trends and Patterns of Activity Regression Techniques for Capacity Planning Weighted Linear Regression—A Preference for the Recent Past Computing the Weights Complex Regression Combinations Forecast Monitoring and Revision Testing the New Model Prediction-Realization Diagram Testing a Working Model with Filter Trips and Tracking Signals Sum of Errors Method Summary of Forecast Monitoring6 Capacity Planning and Systems Optimization Applications Example System Optimization Study for Batch Workloads Background Hardware Environment Software-Initiator Environment Solutions Initiators Job Classes Initiator Assignments of Job Classes Determining the Optimal Number of Initiators Initiator/TCB Time Data Determining the Optimal Number of Job Classes References7 Capacity Planning for IBM MVS Data Storage Systems Overview Disk Capacity Planning Evaluating Current Disk Capacity Evaluating Performance Estimating Future Requirements Evaluating Alternatives Changing Hardware Trend to System-Managed Storage (SMS) Software Available Tape and Tape Drive Capacity Planning Tape Processing—Evaluating Current Capacity Estimating Future Requirements Evaluating Alternatives Software Available Evolving Technologies Data Compression Optical Disks Automated Tape Libraries Redundant Arrays of Independent Disks (RAID) Conclusion8 Charting and Graphics Presentations for Capacity Planning Overview General Suggestions for Successful Data Presentation Labeling the Chart, Data, and Axes Bar Charts Pies, 3-D Charts, and 3-D Pies Charts Specifically Used in Capacity Planning Regression Analysis, Trends, and Curve Fitting Demonstrating Statistical Measures Use of Goal Lines Objects, Symbols, and Information References9 Systems Administration and the Future of Capacity Planning Systems Administration Service-Level Management Problem Management Change Management Performance Management New Directions Systems Isomorphism and Business Elements ReferencesAppendix: A Brief Summary of Soffware Useful for Capacity PlannersGlossaryIndex