As general manager for CA Technologies’ mainframe business, I’ve enjoyed traveling the globe to speak with CIOs and mainframe leaders in dozens of countries. While each business situation is unique, there’s remarkable consistency in how each leader answers the question, “What are the major challenges facing the mainframe platform in your organization?” Hundreds of these conversations, and several quantitative and qualitative surveys, reveal that most responses fall into three major categories: costs, skills and agility. The most common question I hear from these organizations is, “What are other organizations doing to address these challenges?” While the responses to these questions could fill an entire book, the following is a high-level summary of the global challenges and widely adopted solutions.
Challenge #1: Reduce costs. To date, I can’t recall a single conversation where this topic didn’t come up in the first 20 minutes. The phenomenon of the “single, large, integrated bill” puts reducing mainframe costs directly in the crosshairs of the CFO and CIO. An accurate, fully loaded cost comparison of the mainframe vs. distributed platform is useful. The cost challenge must be addressed directly and effectively.
There are two major approaches on the topic of cost:
• “Do more with what you have,” which essentially means maximizing the value of existing software. This involves surveying the mainframe software owned, upgrading to the most current releases, and refreshing staff training to ensure a comprehensive understanding of all functionality. Because of reduced staffing, many organizations don’t understand clearly what they currently own and what the software can do; many discover they’re able to execute new critical initiatives without acquiring additional software.
• “Do more with less.” This sounds daunting initially, but many have and continue to accomplish this with great success. The most common method is to reduce the load on general processors by using specialty processors, including the System z Integrated Information Processor (zIIP), System z Application Assist Processor (zAAP), and Integrated Facility for Linux (IFL). An emerging approach is soft capping and Logical Partition (LPAR) grouping for those customers on Variable Workload License Charges (VWLC) pricing, which requires some time and tuning to maximize the benefit. The highest impact approach is consolidating dozens of software vendors to a strategic few. This approach yields a standardized operating stack worldwide, efficient use of education, training and labor, and ultimately hard cost savings over multiple years.
Challenge #2: Sustain critical skills. The debate rages on regarding the scarcity of available mainframe skills, but the demographics are inescapable. The baby boomers or “the pioneers of IT” are in the latter half of their careers and have started to transition into retirement. Many of these people have 30 or more years of experience with the mainframe, and most organizations don’t have a formal program to capture, retain, and transfer that knowledge from one generation to the next.
There are two essential, related components to the solution. One involves education and formal “critical skills capture and transfer” programs. In the past decade, many university computer science departments have begun teaching mainframe computing courses after a 20-year hiatus. That’s a good start, but the current graduates will require additional training once they enter the workforce. Several leading organizations have developed a mainframe associate engineering program with these attributes:
• Recruiting at universities with mainframe in the computer science curriculum
• A formal internal training program to accelerate general mainframe core knowledge
• A formal mentoring program to accelerate specific mainframe application and business process knowledge transfer
• A formal job rotation program to keep the new hires challenged, engaged, and excited about their new careers.
The second component is adopting modernized mainframe management tools. Modernized tools that have more integration and graphical representation are more relevant to a generation that has grown up around the Internet, gaming, and social computing. With this critical workforce transition occurring over the next five to seven years, it’s time leading organizations update the tools they use to manage the platform in favor of a more modern, contemporary paradigm.
Challenge #3: Increase agility. The final challenge is to increase the agility of the platform to keep pace with the business it’s supporting. For many, agility comes from linear improvement or methodical, ongoing operational improvements. For others, it comes from a step-function improvement or significant surge forward. Today, linear improvements are happening in cross-platform integration, while step-function improvements are under way via cloud computing.
With broad adoption of IBM’s zEnterprise, z196, and emerging adoption of the zEnterprise BladeCenter Extension (zBX) technology, integrated, cross-platform management has become an essential need. A subcategory getting lots of attention is cross-enterprise application performance management—specifically, the ability to ensure Service Level Agreement (SLA) compliance. A problem should be detected and resolved before it impacts the SLA. Fortunately, there are solutions available that integrate best-in-class distributed applications performance management with performance monitoring and fault isolation on the mainframe.
Enabling the mainframe in private or hybrid clouds is the second hottest topic behind cost reduction. While many stalwarts call the mainframe the original private cloud, the question is how to adapt these emerging cloud management technologies to the mainframe to become a more agile private or hybrid cloud. This is an area where technologies are emerging and a step-function breakthrough is on the horizon. The power of the mainframe platform and the massive amount of data it manages and processes is too large, vast, and critical to be excluded from any internal or hybrid cloud. The only questions are when, how, and from whom? Over the next 12 to 18 months, adaptation of distributed cloud-enabling technology to the mainframe platform will likely yield the breakthrough the market is waiting for; this will cement the mainframe’s pivotal role in enterprise computing architectures.
What organizations find most assuring is that 1) they aren’t alone in the challenges they face, 2) others are successfully addressing core mainframe challenges, and 3) they can successfully address them, too.
The only certainties the future holds about the mainframe are that cost pressures will continue to increase, mainframe pioneers will continue to retire, and new computing paradigms will continue to emerge. If the lessons of computing history have taught us anything, it’s that the mainframe will continue to innovate, adapt, and address the challenges that face it, ensuring it retains a vital role in enterprise computing for decades to come.