Many companies outsource their mainframes to specialist outsourcers in order to minimize costs. But the relationship between outsourcer and customer can sometimes become tense due to a lack of transparency—particularly in the area of billing.
Mainframe outsourcing charges tend to be based on MIPS consumption, but there are many ways in which contracts can be framed. For example, you might agree to pay for the average MIPS used per month, the peak MIPS used in the month, the average of the top 10 peaks over the month or some variation of these.
There is additional complexity because MIPS can be calculated in a variety of ways and different organizations use different approaches.
A consultant who advises companies on managing their outsourcing arrangements recently told me about a number of real-life examples that illustrate the confusion surrounding MIPS-based outsourcing charges.
There was the company whose bills were based on an average MIPS usage calculated from the daily weekday peaks throughout the month. The company found out that around 70 percent of all the peak hours used in the calculation were in the early hours of the morning—when the outsourcer’s staff were scheduling all the batch jobs at once. Ideal for the outsourcer’s technical staff but costly for the customer.
Another customer had agreed to be charged in Gartner MIPS and subsequently found out they were being charged in IBM MIPS, meaning they had been charged differently than they should have been for about four years.
Blunders such as this happen not because the outsourcer has set out to deliberately mislead the customer, but because of genuine errors. The outsourcer is usually managing a large, complex infrastructure with numerous staff and many customers and it is impossible to totally eradicate miscommunication and human error. Tensions are compounded because the outsourcer has all the information about the way the systems are being managed and it is difficult for the customer to check if things are being run efficiently and if their bills are accurate.
This is why customers are turning to tools that can help track their MIPS usage and thus provide a check on outsourcers’ charges. These IT business intelligence tools take SMF records that mainframes continuously generate (detailing system operations including I/O, network activity, software usage, error conditions, processor utilization, etc.) and turn this data into useful information such as MIPS usage trends and forecasts or disk storage usage by specific applications.
With this kind of tool, the customer can monitor MIPS consumption rates, identifying the jobs and applications that are producing them. It becomes easier to verify outsourcing charges and to enter into meaningful discussions with the outsourcer if there are areas of uncertainty.
The customer can be alerted to extraordinary peaks in usage, giving them the opportunity to ask the outsourcer to investigate in order to ward off a large bill at the end of the month. This type of analysis should also identify applications that are using CPU inefficiently, as candidates for mainframe application tuning.
Finally, IT business intelligence tools such as these make it possible to map MIPS usage to specific business departments, making it possible to apportion an outsourcer’s charges.
Ideally these tools should be cloud-based so that they do not add to the MIPS consumption and so that the customer can easily access the information and insights they generate via the web.
In any business relationship, tensions and mistrust can emerge if one party holds all the information. So it is in the interests of both the outsourcers and their customers that technology is helping to bring greater transparency into outsourcing arrangements.
Keith Banham has worked in IT for more than 30 years and is the R&D manager at Macro 4, responsible for the company's mainframe suite of products. Keith started as an Assembler programmer at a major bank and during his 29 years at Macro 4 has worked on many of the company’s solutions for application lifecycle management, application performance management, document management and session management. One of his recent roles was the modernization of these solutions by building Web, Eclipse and mobile interfaces, as well as the modernization of Macro 4’s internal mainframe development environments.