Home > Articles > Software Development & Management

This chapter is from the book

Capacity Planning

This process aims to predict future resource requirements and provide a capacity plan for all environments. This process involves gathering and analyzing forecast data from customers, trend analysis of historical data, and workload modeling to predict outcome of growth and upgrades.



Define processes for determining capacity requirements 1–2 years out

Develop equipment plan and associated cost information

Recommend workload balancing options to avoid upgrades

Identify individual components (such as memory and cache) to improve performance and thus eliminate/avoid upgrades

Understanding of key platform, operating-system, and subsystem components

Understanding system measurement data

Basic knowledge of statistical analysis

Understand performance characteristics for all resources

Working knowledge of system/subsystem (such as I/O subsystems) tuning


Automation Technology

Capacity planner

Statistical analyst

Operations research analyst

Capacity planning software (for example, BMC, IBM, Compuware, and others)

Statistical trend analysis

Simulation tools

Analytical modeling tools

Operating-system support utilities

Best Practices


Common organization and management process for all capacity evaluations

Occasional, temporary use of additional resources

Proactive evaluation of all resources

Ongoing tracking of actual versus planned usage by environment and by customer

Tie-in to performance management processes

Service-level attainment

Bulk purchases (for example, on a quarterly basis) for cost-effectiveness

Selective out-tasking of activities to third-party providers

Use of simple business metrics where feasible

Map of capacity requirements onto application infrastructure patterns

IT planned versus actual utilization

Customer forecast versus actual

Number and size of unplanned system acquisitions

Process Integration


Performance management

Capacity on demand

Cross-platform/domain modeling

Integration of service-level and performance processes

  • + Share This
  • 🔖 Save To Your Account