The Reports of Mainframe’s Demise Are Highly Exaggerated
In the world of technology, newer is often seen as better. While this may be true in some areas—the new iPhone 6 is vastly superior to the flip phones we were all carrying around eight years ago—there are some areas where older systems not only continue to work, but are actually better than the new technologies that were built to replace it. We have both worked in the mainframe sector for a long time, and have seen just about every attempt that the so-called experts have made to push the systems to the side so that newer systems can take their place.
Every few years a new argument pops up about why big iron is on its way out, or should be. More than a decade ago, the main argument against mainframes was that they were slow and unreliable. Then empirical data started circulating that proved that large networks powered by System z servers were actually working better than anything else on the market. A few years pass, and the anti-mainframers came back with yet another argument: Running mainframes are much more expensive than clusters of x86-based blade servers.
That all sounded well and good, but again the facts didn’t support the preordained conclusions that mainframe detractors were trying to push. Now in 2014, one of the main arguments against mainframes has nothing to do with performance or price, and instead focuses on the available talent to run the systems. In this narrative mainframe computers are doomed because there are no younger engineers and programmers who have the skills necessary to manage mainframes.
Like the other arguments, it sounds good on paper but falls apart almost immediately. In fact, many young programmers are learning the skills that they need to run mainframe systems, and are landing good jobs early in their careers. So while it’s true that many universities don’t teach specific languages that mainframers need, it is very easy for talented engineers to pick up the tools they need fairly quickly.
That’s why both of us have recently hired several younger mainframe engineers. The process was surprisingly simple: We hired excellent young engineers and took the time to teach them how to run mainframe systems. It exactly how new programmers have always learned to work in the real world, and it seems to be working just fine with this new crop of mainframers.
The anti-mainframe argument hinges on the fact that there aren’t a lot of university graduates who program in COBOL. This may be true, but application modernization takes many forms in leveraging existing COBOL programs in many ways. Ways include re-skinning them with new user interfaces without changing the COBOL code or integrating them with popular languages such as C, C++, PERL, PHP, Java or other languages that just about every programmer knows. And the best part is that these are all available on the mainframe, meaning that the lack of COBOL proficiency does not mean that people can’t work effectively on these systems.
So what does this mean for the future of the mainframe? For starters, it means that yet another prediction of doom will come and go. The technology has been proven time and again, and therefore the arguments against these systems have morphed into a general opinion—at least among those who want to see newer solutions come to the fore—that there aren’t enough people to run the systems because existing mainframe engineers are nearing retirement age. This is—to quote Douglas Adams—pure bunk.
This is all very good news for companies that run on mainframes, as well as for younger engineers who are looking for a particular niche in the technology world. And it all works because many forward-looking companies are taking the time to groom the next generation of mainframe leaders.
We’re not naïve; we know that no matter the facts there will always be people who don’t like mainframes because they see it as old technology rather than as a forward-looking approach to managing massive amounts of data in today’s complex data environment. An infamous American politician once said, “Don’t confuse me with facts, I have a closed mind.” Just remember that the next time someone corners you at a cocktail party and tells you that mainframes are yesterday’s news.
Bryan Smith focuses on IBM mainframe solutions at Rocket Software (www.rocketsoftware.com), a global software company that has developed mainframe tools and solutions for the world’s leading businesses for 24 years.
Daryl Wilson has worked with mainframe systems in IT at a major bank for 30-plus years in Europe, the U.S. and Australia.