How can the mainframe world attract and retain millennial talent? That, and the mainframe skills gap, are two of the more important issues facing mainframers today. We’ll hear from Bill Seubert, IBM Z® Client Architect and mainframe educator, on these topics and more at SHARE Sacramento.
We spoke to Bill on what he has planned for SHARE Sacramento, and to hear about the state of mainframe skills education. If you haven’t already, register for SHARE Sacramento by visiting event.share.org/register.
- Could you describe your background in mainframe? How long have you been involved in SHARE?
I started as a systems programmer, straight out of college in 1984. I worked as an MVS system programmer for three different IBM customers and then joined IBM as a Systems Engineer in 1989. I’ve been with IBM ever since. After my SE days, I threaded in and out of mainframe-related roles in IBM and have been working as an IBM Z Client Architect (and architect manager) since 2004.
As an IBMer, I've worked with probably a hundred or more different companies with Z systems, helping them understand how to use it and reuse assets on Z. I've written Redbooks, given many SHARE presentations and internal IBM presentations and worked with mainframe user groups. Late in 2017, I helped with a blockchain hackathon at a Midwestern University, where the students spun up Linux on Z guests running Hyperledger.
- You’re leading a session on millennials and the mainframe—what do you see as some of the biggest challenges when it comes to attracting millennial talent? How can we overcome those challenges?
In my opinion, there's a stigma about z/OS in some quarters. Many still believe it's old, stale technology—all COBOL and batch, nothing new and interesting. And who wants to work on something obsolete? It's largely a lack of knowledge of what's on IBM Z and its potential. Part of that is on IBM, but part of it is on folks making customer technology decisions based on that incorrect impression. It rubs off on millennials.
How to overcome it? I think there are two ways. First, IBM and sponsors in our mainframe customers must continue to bust the myths and prove that the stereotypes are wrong. Second, companies must begin leveraging new technology on the Z platform, like hybrid cloud, machine learning, DevOps tools and techniques and Linux on Z. For example, put up a prototype of native Node.js and Swift on z/OS, or run an on-premise Blockchain/Hyperledger instance on Linux on Z.
- What are your thoughts on the current state of mainframe skills education? What do you think of the resources available both to current mainframers and to younger programmers looking to join the industry?
I helped write some of the original Z education material for the Academic Initiative, and I've been really happy to see how broadly that material has been used and re-used by various parties to get the skill level built. I'm working with a local university that’s teaching COBOL, and I recently had lunch with a professor from a large university in the Midwest that has an "enterprise systems" curriculum that is largely mainframe-focused. At a higher-ed level, we're really getting better.
Something that really impresses me is the Master the Mainframe competition—it’s doing wonders for changing people's minds on the platform. If a high-school or college student can get up to speed that quickly, you have to start questioning some of the long-held assumptions about how much effort it takes to build skills.
- From your experience teaching these courses, what are the biggest or most frequent roadblocks that mainframe pros run into?
I continue to hear concerns about availability of system resources. After all, anyone can buy a Raspberry Pi for under $50, install Linux and learn Linux on an Intel platform. But it's not that easy for z/OS. IBM is doing things such as the Community Cloud to help, and the Master the Mainframe contest is open to more people now. There are other resources out there, so it's getting better.
The other issue is that there's a roadblock in understanding the "language" of Z, especially some of the more arcane functions like JCL, TSO/ISPF and the other words we use. Some of my IBM colleagues are working on ways to insulate newcomers from needing to know—at least up front—what that stuff is. Eventually, just about anyone working on Z is going to learn about and use those things, but we want to make the ramp-up a lot easier.
- What do you hope attendees walk away with?
I'm trying to help attendees figure out what can be done at their company to make their systems friendlier to mainframe beginners. What kind of tools are available? How can they start leveraging new, "contemporary" technologies on Z that the newcomers might already know from college or other platforms? How can they leverage new cloud and analytics architectures that are more like what you'd find on other platforms?
The SHARE audience is interesting because many of the attendees know the "vintage" Z functionality. So, my intent is to make sure they can relate to the importance of modernizing the platform and infrastructure and not just assume that the newbies have to do it the old fashioned way.
Heading to SHARE Sacramento? Bill will be leading a number of sessions, including “How to Make Your Mainframe Millennial-Friendly,” Wednesday, March 14 from 1:45 pm to 2:45 pm – log in here to save his courses to your SHARE profile. And there’s still time to register for SHARE Sacramento, March 11-16, if you haven’t already.