Service Design - Capacity Management

The official definition of Capacity Management by ITIL is to ensure that cost-justifiable IT capacity in all areas of IT always exists and is matched to the current and future agreed needs of the business, in a timely manner

Some of the processes of Capacity Management include:

• Maintain an up to date capacity plan
• Ensure performance of the service meets or exceeds agreed targets
• Assist with diagnosis resolution and performance
• Use measures to improve performance of services

Capacity management is a balancing act where we attempt to balance cost against resources needed, and, balance supply against demand. We don't need to upgrade to that new server just because a new model has been released.

There are 3 types of Capacity Management:

• Business Capacity Management
• Service Capacity Management
• Component Capacity Management

Capacity Management is kicked off by a variety of monitoring events including:

• Performance events
• Service breach
• New or changed services
• Changes to design or strategy
• Changes to SLA, OLA or contracts

A lot of IT teams start out as they should with component level capacity management, e.g. servers, disk space, CPU usage, storage etc. However, ITIL drives us to strive for understanding business capacity. What are the patterns for business usage and demand? Is the business going to grow, double, etc. If it grows then demand will grow for that particular database application or service. It's critical that we make some kind of estimate or bet over the long run.

In economics, we think of capacity as the relationship between actual output that 'is' produced from our resources and the potential output which 'could' be produced if capacity was fully used. In ITIL, we look at capacity in a similar way. Our primary goal is to ensure that IT capacity meets current and future business requirements in a cost-effective manner.

As the usage of IT Services change and the usage evolves, the amount of processing power, workload etc also changes. We need to understand the demands being made currently, and predict how they will change in the future. The capacity management approach proposes that planning for IT Service growth becomes easier and less reactive.

A common scenario

If there are spikes in processing power at a particular time of the day, you should consider analyzing what is happening at that time and make changes to maximize the existing infrastructure. Changes such as tuning the application for optimization or moving a backup procedure to a quieter our of hours period.

Capacity KPI examples

Reduction in use of old technology causing missed SLAs
Percentage reduction in number of incidents due to poor quality

When we look at Capacity Management we can also consider similar applications such as:

• Application Performance Management
• Network performance management
• Performance analysis
• Performance tuning

Common Capacity Management decisions:

• Which components should we upgrade, memory, servers etc.?
• When should we upgrade?
• Upgrade costs?

At the end of Capacity Management we provide

• Capacity Management Information System CMIS
• The Capacity Plan
• Reports (workload and performance)
• Thresholds, alerts and events

See also:

• Capacity Information Management System (CMIS)
• Capacity Plan
• Performance Reports
• Workload Analysis and Reports
• Thresholds, alerts and events
• Utilization monitoring
• Response time monitoring
• Baselining
• Analytical modelling
• Simulation modelling