Who would have thought that mainframe computers would stay active longer than their caregivers? Throughout the past 15 years, there has been an all-out assault on the viability of mainframe computing vs. network, nee client/server, nee LAN-based computing. Pitched on a myriad of IT panaceas, customers were told to move off the mainframe or face business death, or at least embarrassment at parties. “Can you believe it? Steve said they’re still using the (hand over mouth) mainframe in their shop!” It’s been well-chronicled in this column and others that mainframe computing continues to prove its worth on a variety of fronts despite the predictions. But there is a much larger problem looming that presents the greatest threat of all.
If a mainframe breaks and there’s no one there to see it, does it still stop working?
Even though the rage of the ’80s was lights-out computing, and most people thought HAL would be managing itself by now, the plain fact is computers need people to be their caregivers. The problem is, many mainframe caregivers are not too far from receiving Social Security, planning their sailing adventures, and visiting with the grandkids!
Who is going to care for these critical machines in the future? A recent survey of mainframe technical staff found that 90 percent are age 40 or above, and nearly 49 percent are age 50 or above! That means that for the very systems that serve and protect more than 60 percent of the world’s data, half of their caregivers are a decade or less from choosing another path in life!
Where will that leave the future of the mainframe?
Without a change of course, we’re going to start to see Mainframe Assisted Living centers popping up around the globe. “After today’s lunch we have shuffleboard, Crocheting 101, and z/OS Principles of Operations on the activities menu.”
It’s not just the aging of the mainframe work force that should ring alarm in the hearts of the IT community. The reason that a crisis is on the horizon is there are no troops bringing up the rear to fill their roles. There will be a lot of lonely mainframes. That’s OK; there’s still 40 percent of the world’s data on other servers.
This is a call to action, not just for the mainframe community but the IT community at large. There are some big shoes to fill!
First of all, we have to advocate an all-out war on the tired old interface called the 3270 terminal. It should be permanently retired. Today. Not because it isn’t functional, but because everyone believes that it represents the current state of technology of the mainframe. It simply does not. Even the systems administration interfaces should be rapidly converted to a standard, Web-based interface that the young, pimply-faced techs will be willing to mouse over.
Second, we need to start cross-pollinating! The same survey that revealed the aging populace also showed that more than 80 percent of the respondents have more than 20 years of mainframe experience. Who better to train the up-and-comers than the up-and-came? In order to do that, CxOs around the world need to mandate an immediate détente to the infighting, and start to cross-train systems, applications, and operations skills.
Third, a bridge that provides for current and emerging application skills to easily access and exchange data with mainframes must be built. Web services and XML are providing the means to do that, but we need to put our stake in the ground and do it. This will allow newer generations of programmers to use emerging technologies such as .NET, Java 2 Enterprise Edition (J2EE), Visual Basic, and others to interact with the mainframe.
Finally, universities across the world need to reintroduce mainframe discussions into computing education, from architecture to real examples of its role in the grand scheme. Coupled with entry-level education at companies and organizations, we’ll be able to bring a whole new generation forward.
Computing history has proven that the concept of centralized computing built on a widely scalable, high-reliability platform, a.k.a. the mainframe, continues to serve a critical function. Mainframes aren’t going away—but their caregivers are, rapidly.
Let’s not leave it to the Boy Scouts to come up with a new merit badge to solve the problem: “To obtain his Eagle Scout rank, Johnny had to learn mainframe COBOL, JCL, and how to read a dump, all in six months!”
There’s just too much at stake.
That’s z/Bottom Line. Z