I have spent my career in the mainframe computing space. It’s a strange and insular world. And it’s shrinking a little each day, as the global mainframe workforce sees skilled workers retiring, transitioning to other roles, sometimes even dying, without a sustainable pipeline of mainframe-knowledgeable employees to take their places.
Before starting my current position at a private mainframe software vendor, I worked as a developer for one of the world’s largest mainframe software providers, and later at the IT headquarters of one of the world’s largest mainframe software users. Throughout my time at each, I kept hearing a familiar refrain.
The mainframe is dead! Long live the mainframe!
The story always begins the same way. Big users of Big Iron desperately want to get off the mainframe platform, and once they do, it’s bye-bye Mainframe, and all remnants, real or imagined, of giant rooms of whirring tape reels, blinking lights, and teams of men with 1950s haircuts will disappear forever.
Obviously, that hasn’t happened. The plot is always foiled, playing out something like this.
A mainframe shop undergoes some senior management changes. The new boss, invariably a too-young, MBA-wielding bean counter with no clue what he’s doing, takes a look at the monthly mainframe bill. He recoils into mid-air like a cartoon character, red-faced and suspended in position for a few seconds while smoke comes out of his ears. He then falls back into his executive desk chair and launches an initiative to migrate his company’s operations off of the mainframe.
His underlings earnestly carry that message forward, where it is met with scorn and laughter by the mainframe technical team, a group which, as the story goes, is somehow always made up of older, mostly male employees, crusty and bawdy as they come in a professional workplace. They say they’ve been through all this before, and insist, with a common voice and hacking cough, that the bean-counter is a fool who will eventually come to his senses, just as everyone before him has.
Fast forward about two-and-a-half years, and the migration project has been scrapped, the boss with the MBA has moved on to become CFO at another company, and on the mainframe team, one member has retired, another was laid off, and a third (the youngest and most promising) has transferred to a “safer” area of the company.
Five years later, repeat.
If you are a mainframe-centric person, you know this story well, and can recite the talking points:
- The mainframe platform is not more expensive when all costs are considered.
- The transaction throughput of the mainframe cannot be matched.
- Reliability, availability, scalability, data integrity, and security on par with the mainframe is not available elsewhere at any price.
But mainframe folks also hear the jeers.
- It’s prehistoric technology.
- It’s uncool.
- It’s impossible to learn and use, and no one wants to try.
- Nobody cares.
In response to these criticisms, the mainframe community closes ranks and demonizes those who dare challenge all that is holy. “They don’t understand,” the old-timers lament, before branching into an unhelpful tirade about how an assembler programmer can code circles around some Johnny-come-lately, Python punk.
I don’t believe we devote enough resources to front-ends, dashboards, or even documentation which might help someone who is willing to venture into our world. Once the pioneers, we now often act as members of an elite and entitled club, refusing entry to anyone without the secret handshake and a detailed understanding of first level interrupt handlers.
Yes, it’s a strange and insular world, and even though we say we want and need new blood, people have taken up sides in what increasingly looks like a culture war between the mainframers and the anti-mainframers. Some stakeholders embrace this battle with the partisanship of present-day American politics, with equally unproductive results.
The mainframe has been written off many times, only to emerge time and again as the platform of choice for the world’s largest and most consequential enterprises. In this regard, those crusty old men with their crusty old “I hate Mondays” coffee mugs are right. But the younger generation of programmers are also right. We are keeping them out of the club while simultaneously making it a place where nobody wants to come in anyway. What is the appeal? With all the threats to the mainframe’s existence seen to date, I fear that what could ultimately lead to its permanent demise is that there simply won’t be enough people who know how to work on it.
A former colleague of mine, an excellent mainframe systems programmer, once proclaimed in a meeting that “it takes a special kind of person to do this job.” Was that a boast or a dare? Either way, it’s not music to the ears of anyone hoping to widen the employee pipeline.
Efforts to recruit people to the field, or to insert mainframe courses back into universities, typically involve trying to make the case that mainframes are cool because of what they can do and how important they are.
This approach isn’t fooling anyone. It’s great messaging only for those already on board. Of course it’s true. But once again invoking the analogy of politics, the truth hardly matters. To newly minted programmers with plenty of options, the mainframe path seems both unexciting and too short to sustain a career-long journey. One peek around a bend and they see surprisingly lackluster pay and old men with complaints, instead of stock options and young people with skateboards. Most importantly, they see themselves as having to declare a partisan choice. A binary decision with a big wall separating the two sides. When a new graduate’s friends all go running to the beach party, it’s hard to coax him or her over to the side of green screens and tales of punch cards.
As with politics, we must remember that we are on the same side and that we all want basically the same things. So, before I retire, get laid off, or die myself, here are a few suggestions to help replenish the mainframe workforce and save the world.
Let’s show how the client-server side of a business works in conjunction with the mainframe side, not separately from it or against it. And show the synergy the web and social-media departments have with both. There is already a lot of common ground. Metal-C, Mainframe-based Linux, AI in all flavors. IoT needs both mobile devices and mainframes to operate.
Instead of apologizing for areas of overlap or highlighting them as potentially desirable exceptions, we should champion the wide open, undivided playing field that is modern computing, until such thinking becomes the status quo.
In all its forms, we must tear down this wall.
David Herlich is a Senior Mainframe Software Developer who most recently worked at DataKinetics—the global leader in Data Performance and Optimization. He started his career at IBM, where he spent over 25 years in the Db2 for z/OS development organization as a developer of the product and as an evangelist for the platform. He earned his degree in Computer Science and Mathematics from the University of California, Davis, along with a double-minor in Communications and Psychology.