Popular Methods Used to Determine the Capacity of Mainframes

Managing the capacity of mainframe computers plays an important role in controlling the system expense of any company. Decreasing CPU/MIPS/MSU consumption can lower costs. Many of IBM and ISV software licensing charges are connected to peak capacity or to a rolling four-hour average of consumed processing capacity. Today determining the capacity of a mainframe is done several different ways.

Managing the capacity of mainframe computers plays an important role in controlling the system expense of any company. Decreasing CPU/MIPS/MSU consumption can lower costs. Many of IBM and ISV software licensing charges are connected to peak capacity or to a rolling four-hour average of consumed processing capacity. Today determining the capacity of a mainframe is done several different ways.

MIPS
MIPS are usually defined as Millions of Instructions Per Second or a calculated number used by many system managers to determine the capacity of a mainframe computer. However despite its wide spread usage, many analysts believe that a MIPS number does not really provide those managers with a useful measurement of mainframe capacity. MIPS measures processor speed alone. Because of this, MIP numbers do not include a number of variables like processor speed, speed of input/output devices, memory speed, or types of workload.

MSU
MSU (Millions of Service Units per Hour) is generally used to calculate the cost of software licensing. Vendors generally base their software charges on an MSU rating. Charges are based on a MSU rating that is derived based on millions of central processing unit (CPU) service units per hour. MSU values were developed to provide a base number for vendors to use to charge for software usage across multiple processors.

ITR
IBM recommends that all its customers use the Internal Transaction Rate (ITR) method to measure mainframe capacity. ITR numbers are derived by using workload type to ascertain the capacity of a particular mainframe using number of transactions per CPU second. This enables system managers to measure capacity based on different workload configurations. Using this method, systems manager can measure batch, online transactions as well as web activity. The important thing is that IBM uses the average ITR for calculating MIPS values.

Summary
David Stephens of Longpela Consulting Capacity recommends that ITR should be used for serious capacity planning, MIPS to get an easy number to use for planning and MSU to determine software cost. An accurate measure of a mainframe’s capacity is important because it gives planning analyst important information.

With accurate system capacity information a planning analyst can warn the systems manager if a system has stopped effectively servicing user requests, track workloads and proactively recommend system updates. Determining a mainframe’s capacity determines IT expenses which directly affects an enterprise’s bottom line. Capacity planning is effective only if it identifies system problems before they happen and enables a company to reduce IT cost.

Since 1982, Software Diversified Services is providing first-quality software and technical support for IBM mainframes and VM, VSE, mainframe ftp and z/OS systems. SDS mainframe software provides security, encryption, and compression for mainframe data; secure file transfer (SFTP); SNA and IP network management and automation; and performance monitoring for applications and networks.

License: You have permission to republish this article in any format, even commercially, but you must keep all links intact. Attribution required.