When it comes to mainframe computing, business IT planners and practitioners alike confront a significant challenge that will require an industrywide effort to rectify. These planners have an affirmative view of the mainframe. They agree that it persists as a core platform for corporate computing. Its performance characteristics and cost of ownership advantages over alternative platforms are understood and appreciated. From the standpoint of cost, compliance, continuity and carbon footprint (utility power consumption), the business value case for the mainframe is rock solid and is even underscored by both current economic realities and the prevailing regulatory climate.
Yet, I’m being told that plans to leverage this technology over the longer term are challenged by an inescapable fact: the aging mainframer. To planners, staffing issues loom as a significant risk for future mainframe computing, partly because enrollment in IT-related disciplines is generally well below past levels, but also because only a fraction of those who are developing the necessary skills to support information technology are choosing mainframe technology as a career path.
In truth, the future mainframe skills shortage is already being felt in organizations today. As mainframe staffers retire, remaining staff are being tasked to perform the work once distributed across a hierarchy of well-defined roles. Rather than specializing in specific functional areas, the modern mainframer is becoming a jack-of-all-trades, required to interact with a variety of subsystems, application software, and management components to perform virtually any task.
The blurring of traditional mainframe data center staff roles is beginning to inhibit the legendary efficiencies of mainframe computing. A database manager recently told me that he must now take direct responsibility not only for his transaction processing software, but also for the allocation and management of machine resources. Rather than passing these tasks on to domain experts responsible for networks, storage and machine resources, Mr. or Ms. Database Manager must perform this work directly—typically by leveraging a variety of specialty software tools intended by their vendors to simplify (reduce the knowledge prerequisites for performing) such work.
Given the new workload being migrated to the mainframe platform in many shops, and the proliferation of software tools that all mainframers must master to do more with fewer workers, it’s no surprise that troubleshooting and break/fix work are becoming the big “time vampires.” Seasoned mainframers have less time for “core activities” such as application development and planning that drive real information technology innovation and business value. Moreover, there’s little time to do what must be done to address future mainframe staffing requirements: on-the-job training and mentoring.
The old adage holds that it takes a decade to create a mainframer. In the past, someone choosing this career path proceeded through a set of well-delineated roles and responsibilities, starting as an operator or tape librarian, then moving into application or resource management domains, and ultimately into system architecture or programming or management functions. Specialized training combined with hands-on experience cultivated knowledge and skills that ultimately enabled the staffer to become a domain expert for life (a “virtuoso”) or, in a few cases, a cross-domain expert who could serve as a senior planner or manager for the overall plant (a “maestro”). The well-defined separation of skills areas in mainframe computing aided the orderly transfer of knowledge from one generation of mainframers to its successor. Perhaps more important, mentoring played a significant role in knowledge transfer as a key component of on-the-job skills development.
Given the current “blurring” of domain responsibilities and the consolidation of workload onto fewer workers, time for mentoring is in extraordinarily short supply. The truth is, even if colleges and universities were producing droves of mainframe careerists, there’s no time to groom them to fill tomorrow’s data center operations or management roles.
The benefits of mainframe computing are inextricably tied to the ongoing development of domain virtuosos and cross-domain maestros who make an information processing symphony happen from a collection of diverse hardware and software instruments. For good reason, business IT planners are questioning whether they will be able to keep the system viable and if so, for how much longer. The questions are sobering and real, and they need to be addressed with less hype and more action.
It’s time for the industry to help address these issues. We need to throw some real technology at updating the mainframe experience; this includes how it maps to changing work models, how knowledge can be transferred to newbies, and how we can integrate the myriad element management products into a more holistic and task-oriented paradigm.
I’m encouraged by what I’m seeing at CA Technologies and elsewhere. You owe it to yourself to become familiar with what’s going on in these companies and to tell your vendors you want them to pursue similar efforts to integrate and optimize the mainframe experience. That’s the only way we can keep the often projected demise of the mainframe from becoming a reality.