Melpomene - Fotolia

Former FINRA CTO: Writing's on the wall for IT operations careers

Those who aspire to an IT operations career must learn to code, says a former FINRA CTO who works with financial institutions as a DevOps consultant.

IT operations careers are in jeopardy unless IT pros quickly learn to code, even at conservative businesses such as financial services companies.

That's the view of Marty Colburn, the former CTO of the Financial Industry Regulatory Authority (FINRA), and former CTO and managing partner at a cloud and DevOps consulting firm. He now works with a who's who of big financial firms, such as the Federal Reserve and American Express, to transition them to a cloud-based, DevOps-oriented IT workflow.'s Beth Pariseau caught up with him to get the lowdown on what's happening with IT operations career paths.

What's been the effect on IT operations careers at financial institutions during these DevOps transitions?

Marty Colburn: The only IT operations folks that I've seen make the switch are folks that go out and learn development languages, and implement them in some cases where you have more procedural tools. In large part I've seen them put into a very difficult position because their skill set starts to become obsolete very quickly. You need people [who] can script, move fast and understand the technologies in a much more innovative way.

What tools do IT ops pros use who've made the leap?

Colburn: At least on the procedural side you see more of the IT operations folks move to Puppet. When they start to get into Chef or Ansible or any of the scripting, even the Maven and the Jenkins for continuous deliveries, you see them have to retool. I haven't seen many of the operations folks I've worked with make a concerted effort to retool. It's almost like they hunker down and hope this phase or phenomenon will move along and they'll continue to run operations the way they've traditionally run it.

So when you say they have to retool, what tools are they using?

Marty Colburn, former CTO, FINRAMarty Colburn

Colburn: My first counsel is to start with the basic concepts of object-oriented programming, to start looking at things like abstraction, inheritance, encapsulation and then start looking at the development languages that allow you to start adopting the DevOps mindset and determine where you want to focus. For example, if you're going to build in Selenium, you have to have a Java background. You start with those object-oriented concepts, then you learn Java and then you can actually write on that framework, and that becomes one of the [key pieces] of automated testing. And then as you think about the continuous delivery and integration, you start to focus on the automated build aspects.

Once you start moving into a DevOps environment you're coding, similar to application developers, but you're coding for different things. So writing a script to do an automated build is not fundamentally different from running a Python script to build a data grid. You have to have that kind of building-block foundation of software engineering to be effective in this space. For years infrastructure engineers didn't have to have that background, and that's changing rapidly.

So the writing's on the wall so to speak for IT ops pros.

Colburn: It is, absolutely.

What else can IT ops people do, besides learn how to code, to keep up with this transition?

You have to have that kind of building-block foundation of software engineering ... For years infrastructure engineers didn't have to have that background, and that's changing rapidly.
Marty Colburnformer CTO, FINRA

Colburn: When I look at folks that have to retool I start with just an Agile methodology. When you think about technical project management, at least have the aspect of how to run scrums, in a safe methodology or a scrum of scrums and a release train, and how to take on traditional project variables -- cost, schedule, scope and risk -- and be able to incorporate them into a rapidly paced environment. And from there, it starts to dictate the tool sets you have to develop and learn.

The ones that I've seen not move forward try and fight DevOps and automation and an ability to modernize their environment. You have to do it in a very rapid fashion and the ones that take a long time setting those goals don't make a lot of progress.

What about the site reliability engineer model? Does that represent an opportunity for IT ops?

Colburn: Absolutely. That's becoming the nomenclature for what I would call DevOps 2.0. That's become one of the fastest growing skills that start to get IT operations folks moving to it. It starts to tie together not only DevOps but also security with it.

Beth Pariseau is senior news writer for TechTarget's Data Center and Virtualization Media Group. Write to her at [email protected] or follow @PariseauTT on Twitter.

Next Steps

DevOps requires both sides evolve to come together

You can code your own DevOps infrastructure

AWS has enough code options for Dev and Ops to collaborate

Dig Deeper on IT Ops Implications of Continuous Delivery