Capacity Plan

Capacity planning is the process of correctly sizing a system to meet specified performance objectives.  Capacity is taken to mean the amount of work that a server is capable of completing in a given period of time.  In a simple model, it might be calculated as (processor group) x (number of processors) x (application) x (memory).
 
The demand a server experiences will vary under many circumstances.  Notable events that might cause the demand for work to vary greatly include number of users accessing the application, configuration of system values, amount of hard disk space, allocation of memory, CPU utilisation etc.
 
Discrepancy between capacity of a server and the demands of its users results in an inefficiency, either in under-utilised resources or unfulfilled user experience.  The goal of capacity planning is to therefore minimise this discrepancy.
 
Upon engaging PCS, we will provide a Capacity Planning Guide document to gather relevant data.  For each server subject to this service we require:
  • iSeries performance monitor data for the creation of workloads;
  • A list of performance and throughput objectives for each workload and/or user group;
  • Existing workloads, where possible; and

 
After receiving the appropriate information, PCS will:
  • Build a capacity planning workload for each user group or application identified;
  • Use one or more capacity planning tools and/or other tools to match requirements to an appropriate iSeries configuration; and
  • Present a written report of the capacity planning results and recommended configurations.

    Click here for a Capacity Plan sample in PDF format.
For more information on how we can assist in this area click here.