There has recently been a lot of attention to a number of demographic trends that may impact the mainframe world. One, of course, is the impending retirement of a large portion of the mainframe workforce. Another is the retirement of the baby boomer generation, which will be proceeding for the next 20 or so years. The third is declining IT enrolment in colleges and universities.
You'd think, with all of those factors, that we'd be heading for a perfect storm on the mainframe - and from a certain perspective, you'd be right. We are heading for it... but the good news is, not only do we have a choice about how we deal with it, but in fact it may save some of the largest organizations on earth from an even more devastating storm.
That bigger storm is all the rest of the world's accelerating IT staffing needs, coupled with a comparatively shrinking workforce.
The total number of distributed servers and their power and storage continue to increase, along with the interconnected nature of their applications. This means that the staffing needs for them are increasing faster and faster. But there aren't even enough people to handle today's demands, and the number is not growing at anything like the pace of the need.
Of course, solutions to unify and simplify the management of IT enterprise-wide are an important response to this growing complexity and scale, and they'll certainly take some of the load off. But you still need people in the equation.
So, we find ourselves in a world where the mainframe is about to run drastically short of staff in the short term, and the rest of IT is going to run out of staff compared to accelerating capacity and demand in the medium term: looks like we're all going to be up a pretty big creek without a paddle, doesn't it?
Here's where I get excited as a mainframer: instead of being the crisis that portends an even bigger crisis, the mainframe can actually become the solution that forestalls the bigger crisis, at least for those organizations that have (or acquire) mainframes.
And that's because the mainframe problem is eminently solvable.
After all, even though the amount of mainframe capacity has grown substantially over the past few years, the number of people managing it has generally stayed the same or gone down, thanks in part to better and easier-to-use systems management software and other advances.
When the mainframer retirements begin to reach a visible level, whether or not organizations have planned to deal with them, they'll have to start. While there may be a few bumps in the road, most companies will find a way through, and within a few more years we should have a healthy, fully-staffed mainframer population. Solution achieved.
Once the mainframe staffing problem has been solved, be it sooner or later, it will be increasingly apparent that this is the one platform that has none of the issues that are already challenging distributed systems, and only look to get worse. I suggest that the response will be to wake up to the amazing quality and ROI of the mainframe and begin moving relevant applications there in order to make them functional, reliable and manageable with an acceptably finite number of staff.
I'm not suggesting that we're going to see an end to distributed computing - it's too ubiquitous. But I do think that the largest organizations on earth - i.e. those big enough to have mainframes - will have the potential to rise above many of the staffing and complexity problems that the distributed world is going to be dealing with for the foreseeable future.
And that's a nice problem to solve!