An overview of techniques available to address capacity planning in the production data processing environment is presented. The production data processing system is briefly described and its capacity is quantified. The measurement tools, reports, and data required to implement a capacity planning program are discussed. Modeling and prediction are placed in perspective with the overall objectives of the capacity planning process. Personnel (managerial and technical) and organization considerations are also discussed.

Overview of the capacity planning process for production data processing

The performance of MVS (Multiple Virtual Storage) systems can be predicted for changes in workload and environment by an IBM marketing aid informally called SCAPE (for System Capacity and Performance Evaluation). Written in FORTRAN, the programs use simple queuing formulas with empirical modifications. Response times for complex workload(IMS, CICS, TSO, and batch) through the CPU and auxiliary storage are expressed as functions of application loads and other parameters that define the system's environment. SCAPE can predict the effect on performance of different CPU models, larger memory, additional channels, additional direct-access storage, larger block sizes, and alternate workload projections.

System capacity and performance evaluation

This paper extends a particular capacity planning approach to include usage accounting by business element of Input/output path and Direct Access Storage Device resources. A simple nonlinear procedure is outlined to size host configurations that can process workloads at specified rates. An important feature of the procedure is to take account of a law of diminishing returns, which is that doubling the number of components does not double the amount of work done. Discussed are configuration relationships involving TSO and DB/DC subsystem sizing, tuning, workload variability, data considerations, and hardware and software considerations. Typical, but hypothetical, examples are presented.

Processor, I/O path, and DASD configuration capacity

This article defines computer capacity planning and proposes a four-phased model for computer capacity planning (CCP). The model contains the major steps needed for developing a comprehensive computer capacity plan. The requirements for capacity planning are defined in terms of people, hardware, software, and computing resources. Common tools and methodologies for each step in the planning process are described, including a discussion of the advantages and disadvantages of each. This article provides guidelines for the development of a formal capacity planning function.

Computer capacity planning: strategy and methodologies

This thesis describes the development of a new capacity planning methodology to aid computer professionals working on implementing a capacity planning effort in an MVS environment. The methodology developed utilizes the System Management Facility (SMF), the Resource Measurement Facility "5 (RMF), and the Statistical Analysis System (SAS) in order to bring together many of the past computer performance evaluation ideas and to integrate them with current computer performance evaluation techniques and automated procedures. An automated approach is taken so computer professionals can devote their energies to analyzing and interpreting results, not gathering and calculating input factors. Subsequently, each MVS installation is supplied with the actual usage data to help set accurate user objectives and forecast for tomorrow's data processing needs. Accesion For NTIS CRA&I OTIC TAB alhflnnotn;Ced C-.. ...........................

An Introduction to Implementing a Capacity Planning Effort in an MVS Environment.

Discussed is a capacity planning procedure called USAGE. Various business elements to be individually measured and tracked are presented. Outlined are methods for estimating workload growth. Separate limits of capacity for on-line workload and batch workload demand are discussed. A simple graphic presentation procedure is included to communicate the results of a study to those who need the information for making business decisions.

J. C. Cooper

Papers

A capacity planning methodology