Sergey Nivens - Fotolia

Curiosity didn't kill the mainframe systems programmer

Mainframe systems programmer Robert Crawford explains the lasting relevance of mainframes, his mixed feelings on certifications and how curiosity continues to help him.

Many people consider mainframes the Ferrari of server workload performance. Or at least Robert Crawford does.

Crawford, a mainframe systems programmer with over 34 years of experience, specializes in CICS technical support, as well as VSAM, DB2, IMS and other assorted mainframe products. He has programmed in assembler, Rexx, C, C++, PL/1 and COBOL.

After getting his Bachelor of Science in computer science from Texas A&M University, Crawford worked for three years as a CICS systems programmer. He then joined a large insurance company as an operations architect responsible for establishing mainframe strategy and direction, and has worked there for close to 31 years. His job and skill set has evolved as mainframes evolved over 30 years.

Crawford shares his expertise with SearchDataCenter as a contributor and advisory board member.

What advice would you give someone just starting out?

Robert Crawford

Robert Crawford: If you're new to the mainframe arena, I would advise you to spend some time delving into some of the more technical subjects like assembler language, debugging, system internals and performance. These skills have fallen by [the] wayside for many years, and I see too many people becoming system janitors instead of technicians. Others may argue the modern mainframe systems programmer doesn't necessarily need to know the ones and zeros, but I think losing these skills leaves shops at a disadvantage and at the greater mercy of IBM. Also, learning low-level technical stuff will make you more valuable as these skills become rarer.

Do you have any technical certifications? Are they necessary?

Crawford: I don't have any technical certifications. To be honest, I have mixed feelings about certifications. On one hand, I know they take a lot of work and some of them can be quite valuable. On the other, sometimes the certificate just means someone learned what a particular vendor thinks is important.

What is the most challenging issue you have ever faced in the data center? How was it solved?

Crawford: Probably the most challenging issue is one I'm working on now, active/active. Not only is setting up the individual A/A software components difficult, the design of resilient applications and planning day to day operations is a whole rat's nest of problems. There are so many compromises and duct tape fixes involved -- it drives me crazy.

What is the most important lesson you have learned working in the data center?

Crawford: The most important lesson I've learned is to follow my curiosity. Sometimes, while doing what I'm supposed to be doing, I find an interesting thread I can follow. While following the thread isn't exactly part of my immediate job description, it leads to something new that's good to know. These investments in time don't always turn out to be useful, but they do make my job more rewarding.

How do you see mainframes evolving? Or rather, how do you see mainframes staying relevant?

Crawford: Mainframes are and always will be relevant. Frankly, the idea that mainframes need to redeem themselves rankles [me]. All platforms have their strengths, weakness and quirks, but the mainframe is very good as what it does and represents the state of the art in many respects.

Frankly, the idea that mainframes need to redeem themselves rankles [me].
Robert Crawford

To be honest, the reason mainframes have lasted so long is inertia and IBM kind of counts on it. They understand people have been running mainframes for a long time and it is very difficult and expensive to get off of them. On the other hand, IBM has done quite a bit of investment in hardware and software to keep the mainframe capable and state of the art. When you're talking about security, throughput and managing disparate workloads, mainframes are very good at those tasks, while also handling modern tasks like Web services and Java.

The future mainframe will continue to support the oldest and newest applications very efficiently. From an administrative and development point of view, the mainframe will develop GUI interfaces that will boost productivity and be familiar to college students entering the workplace.

Where do you see the industry heading?

Crawford: I'm not sure where the IT industry is heading. What's amazing after 35 years of doing this is how history repeats itself. The 80's began with monolithic mainframes attached to dumb terminals. The 90's brought us distributed systems, PC's and smart terminals. Now we see thin clients, lightweight communication protocols and data center consolidation. Whatever comes around the corner next is likely to be a recapitulation of something else.

What is the worst IT question a friend or family member has asked you?

Crawford: For the most part, my friends and family don't have a lot of questions. I do, however, have a tough time convincing them that computers do exactly what you tell them to do. Therefore, if a file is missing or a document badly formatted, it's likely something they did as opposed to malice on the part of the computer.

Next Steps

Prepare to ace an intermediate-level COBOL interview

Generate and maintain quality mainframe test data

Four ways to beat a mainframe skills shortage

Understand SIMD instructions on the z13 mainframe

Dig Deeper on IBM system z and mainframe systems