Whites Papers

The traditional IT management measurement system serves two important functions. It provides the current status of a problem and it indicates in which direction that problem is trending. For the purpose of this document we will refer to that measurement system as a one tier system. A one tier measurement system is an internal informational system.

The first tier of a three tier system performs the same functions as a one tier system, but because that first tier is linked to its second and third tiers, the three tier system has a far greater informational value to IT management. This three tier system also has the added advantage of causing most measured problems to be resolved by the vendors whose products and services are often the cause of those problems.

R+2 is an example of a three tier system that is used to measure the stability of the corporate network system.

Tier One

Each day the first tier of RPI's three tier system (R+2) gathers the count of all network server machines operating on the network and the count of the interruptions caused by each of those machines. Each month from those daily counts the Interruption Rate and the Uptime Percentage rate of the network are computed. These monthly rates enable IT management to track the trend of the stability of its network. In addition to the monthly network stability rates, six additional sets of rates are produced that are used to increase IT management’s understanding of which components on the network produce the higher levels of interruptions. Those additional rates are computed for the two network server platform options (mainframe and traditional servers), for the several operating systems that support the machines on these two platforms, for the several hypervisors that facilitate virtualization, for the individual vendor’s hardware that run those machines, for the hardware maintenance providers that maintain that hardware and for the two operational states, virtual and non-virtual.

Tier Two

The second tier of RPI’s three tier system compares those internal stability rates to industry standards to provide added context and to give guidance on which products and services IT management should choose in order to maximize the stability of the network. Measuring the current stability of the network and knowing if the stability is increasing or decreasing is important, but knowing how that stability compares to the average stability in the industry, knowing how it compares to the best stability in the industry and knowing where that stability ranks in the industry are also important. The first tier tells management where its stability is and in what direction it is trending. The second tier tells management where its stability should be and where it could be. A one tier system uses the IT organization as its own standard of what is normal. The second tier of a three tier system uses the industry as a whole as the measure of what is normal and what is possible.

Tier Three

The third tier promotes competition between those vendors whose products and services are the source of most of the interruptions that undermine the stability of the network. The third tier is predicated on the concept that those that create the problems on the network and have the resources to solve those problems should be provided with access to the measurements that clearly show their culpability for those problems. A three tier system is based on the concept that the best measurement system is one that not only has the power to inform management of the status of a problem, but also has the power to resolve the problem. Each month RPI sends a report to each vendor organizations that provide products and services that can impact the stability of the corporate networks. Included in that report are the comparative interruption rates of that vendor’s products and services and the rates of its competitors. History tells us that the vendor with the best rates will see to it that those rates are presented in every sales situation where its products or services are being considered and will do everything it can to keep those rates higher than its competitors. Conversely the vendors with less than the best rates will either have to accept that their market share will decline significantly or their rates will have to increase significantly. Ignoring the rates is not an option. It is not the good will or the good intentions of the vendors that will provide an interruption free network, but their desire to remain competitive in the marketplace. The third tier is designed to take what has up to now been viewed as an IT management problem (interruptions of the network) and transform it into a vendor management problem (loss of market share).     

A three tier measuring system is not a theory

The original R+ measurement system introduced in 1970 and eventually used for about thirty years by more than two thousand IT data centers in the western world was the industry’s first experience with a three tier measurement system. Tier one measured the reliability of mainframe hardware in each of the participating IT organizations. Tier two compared those measurements to the average and to the best reliability in the industry. It then ranked each individual data center’s hardware based on its reliability by model and vendor. Tier three provided each hardware vendor organization the result of those monthly measurements. Some vendor organizations could not sufficiently increase their reliability and were forced out of the marketplace. The remaining vendors competed over the next three decades to be the best and as a result of that competition the problem R+ was measuring virtually disappeared (see Remembering R+ and what it accomplished).