Lower your mainframe MLC costs and fund IT innovation projects instead.
It’s such an exciting time to be in IT. New technologies and usage models demand that companies transform the way they deliver services to drive success in the digital enterprise. Think about it. Social platforms expand collaboration. Mobility provides unprecedented flexibility. Cloud computing increases scalability, agility, and much more.
Yet when people look at IT budgets, particularly to pay for innovation and digital business transformation, they can’t help but carefully examine the total cost of running a mainframe environment. They must be careful, however, because the wrong approach can undermine the ability to deliver services. The mainframe is often viewed as one of the largest overall expenses in an IT budget because it usually appears as a single line item, and so it’s frequently a target when organizations are looking to fund other strategic business initiatives.
In reality, the mainframe is actually less expensive to run than some other platforms that split the bill into smaller parts. The answer to the mainframe budget challenge isn’t to go after the mainframe itself, but rather to reduce the mainframe’s monthly license charge (MLC) software costs while ensuring service delivery. There are a variety of ways to make this happen.
How to Get Costs Under Control
Did you know that the greatest cost driver for the mainframe, representing 30 percent or more of the total budget, is typically the MLC software costs? The key is to get this expense under control, especially as IBM® MLC prices continue to increase by about four to seven percent a year while mainframe budgets stay the same or even shrink. Fortunately, with better visibility, predictability, and automation, you can lower these costs by up to twenty percent or more. These actions can enable your organization to redirect funds to high-value innovative projects without impacting the existing service delivery you expect from the mainframe.
Here are five proven strategies – based on real customer results – to consider for reducing your MLC costs.
- Identify cost drivers and pull non-time-critical workloads out of peak periods. IT should understand usage profiles that are driving the peak 4-hour rolling average (4HRA), the workloads that are driving the high MLC costs, and when they can be moved to a lower utilization period. Real world example: An organization rescheduled work to start an hour earlier, based on information provided by a cost analyzer solution, which removed processing from peak times and saved over $100,000 per year in MLC costs.
- Model the cost impact and consequences of changes before making them. With predictive modeling, IT can anticipate the impact of change on the peaks and how that translates to changes in MLC costs. Real world example: By modeling MLC costs with a cost analyzer solution, and using the results to gain consensus among applications development, operations, systems, and the business owner, an IT organization moved some IBM DB2 work from one logical partition (LPAR) to another and saved more than $150,000 a year.
- Use dynamic, intelligent capping to limit the peaks without impacting the business. With an intelligent capping solution, you can avoid the problems that come with manual approaches to capping while ensuring that high-priority business work runs without constraints. Real world example: One organization saved approximately $40,000 a month in MLC costs with intelligent capping and mitigated the delays of important business work.
- Optimize where subsystem instances are running to reduce subsystem license costs. When you optimize how and where subsystems are used, IT can change the math of MLC software to lower your MLC costs. Real world example: By consolidating expensive MLC products onto fewer LPARs, one organization saved more than $300,000 a year in MLC costs.
- Tune the work that runs during peak periods. While it’s important to effectively allocate workloads and subsystems across the infrastructure and time, you should also consider how to optimize the work that makes up the peaks. This approach can reduce CPU usage of major business applications. In addition, by tuning the management software that also runs during peak times, such as real-time monitors, or choosing more efficient mainframe monitors, you can further reduce monitoring CPU usage significantly. Real world example: An IT organization discovered that the major driver of the 4HRA peak was an online business application. After using application tuning technology, this group reduced requirements during peak by 35 percent and improved service level performance by 30 percent. By reducing consumption, they were able to postpone a planned CPU upgrade and substantially reduce the MLC bill.
By using a combination of the five techniques discussed here, you can enable IT to reduce MLC costs dramatically. The actual savings will depend upon the applications, MLC software products, and usage patterns of the environment. The cost reductions you’ll achieve can help fund digital transformation and maintain service quality, a worthy outcome for your efforts.
- Mainframe MIPS: An Introduction
- Why It’s Good to Grow your Mainframe – Key Insights from the 2016 Mainframe Survey
- Your Mainframe Usage Should Not be kept a “SCRT”