Looking Ahead to the Next Decade of DevOps
Technologist and CA Senior Strategist Peter Waterhouse says data science and self-modifying code are in DevOps' future.
Predicting the future can be a fool’s errand, but that’s never stopped anyone from breaking out the crystal balls.
For best results, we look to clear-eyed and far-sighted experts like Peter Waterhouse, Senior Strategist at CA Technologies. With more than 20 years experience on the cutting edge of business technology, Waterhouse is a noted blogger and speaker on topics like mobility, cloud, IoT—and especially DevOps, which entered IT lingo 10 years ago.
The biggest change to expect in DevOps during the next decade, as Waterhouse sees it?
“Business conditions are changing so quickly that it requires IT to provide a more rapid, instantaneous response to new triggers,” he says. “Immediacy will become a necessity, particularly in 10 years when IoT will be proven at scale and nascent technologies like serverless and quantum computing and fifth-generation networks become prevalent.”
Below, Waterhouse shares insights on the coming decade’s challenges and opportunities for software development.
Modern Software Factory Hub: DevOps has succeeded in getting code out more frequently and much faster via continuous integration, testing and delivery. Have these practices reached maturity, or will they continue to evolve?
Peter Waterhouse: Continuous deployment is part of the movement to the shift left principle, where you take disciplines like performance monitoring and security assessments and move them to earlier in the development cycle. What’s been missing in DevOps is the notion of actually improving the health of applications as they progress rapidly. As more businesses adopt the shift left approach, they will end up not just with developers providing rich applications that get out even faster, but with applications that are of high quality, too—reliable, resilient, supportable, scalable. That said, no matter how much coding dexterity developers build up, they’re never going to keep up with the pace of change that businesses will increasingly be living with. So, we’re going to also see advances move beyond what developers currently are doing to anticipate change, such as incorporating feature flags in their applications to toggle functionality on and off, and using self-learning, self-modifying code driven by artificial intelligence and cognitive computing.
I think big data and analytics is where DevOps can really shine.
— Peter Waterhouse, Senior Strategist, CA Technologies
MSF Hub: During the next 10 years, how do you envision the relationship between DevOps and big data changing?
PW: I think big data and analytics is where DevOps can really shine. Even though businesses today may consider themselves data-centric, many still struggle with understanding how to get value out of mountains of data. That’s because the work being done by their skilled data scientists is typically isolated from mainstream IT and DevOps. So, while data scientists do a great job finding the golden signals in data and even codify them, they lack access to production IT systems for testing. This means their hypothesis only gets verified when their production goes live. And if it goes live and the hypothesis is wrong, you have a big problem.
But if we look at data scientists as “new developers,” they can start to work collaboratively with ops to really drive the value of their solutions. Again, it’s a sort of shift left scenario, where data scientists can prove or disprove their hypothesis and analytics before production—the same way that traditional developers verify their codes before production. DevOps isn’t on most people’s radars as applying to data science and big data yet, but it will get there, because those are the apps that will truly differentiate a business.
MSF Hub: How will businesses have to change the ways in which they leverage DevOps to truly enable digital transformation?
PW: DevOps must scale to support digital transformation, but there’s a disconnect when businesses think of digital transformation as doing the same thing they always have done in the same way, only faster and cheaper. They wind up using DevOps purely as a means to reduce costs and increase current efficiencies.
The winners will be those who leverage DevOps to change their operating models and support strategic pivots. Think of Netflix as an example of conducting an existing profitable business in new ways: The company didn’t fundamentally alter what it did, but it changed its operating model from shipping DVDs to streaming content.
This all goes hand in hand with framing any DevOps initiative in the context of the customer experience and what that will mean to revenue, profitability and margins.
Instead of monitoring systems in network operations centers, for instance, [IT ops] will morph into digital craftsmen.
— Peter Waterhouse, Senior Strategist, CA Technologies
MSF Hub: Do you have any predictions on how developers’ roles may evolve in the coming years?
PW: On the developer end, low code will have an impact on the work that now is done manually for addressing issues like security, cross-platform support and data integration. That, along with self-learning and self-modifying code, will change developers’ practices, hopefully helping them to focus on business problems more than on technical requirements.
In the next 10 years, IT Ops will act as a business itself. These teams will build and market competitive services that other teams in the organization can leverage. Instead of monitoring systems in network operations centers, for instance, they will morph into digital craftsmen.
Think of the idea of a software factory, where the folks who work across the factory are the ones who drive improvements—that’s what developers will do. They’ll work across the application lifecycle to improve performance. It’s part of the focus shifting from an end-of-cycle, cost-centric view of ops to it being much more proactive across the digital delivery chain.
We’ll also see self-organizing DevOps teams becoming key to a shift in how IT manages itself, going from being aligned around technical functions to being aligned around the products they deliver to customers.
MSF Hub: How do you expect C-suite leadership’s perspective to evolve regarding the value of DevOps to its business success?
PW: The true test of DevOps is whether it can be openly measured and discussed in the C-suite boardroom as bringing intrinsic value to the business, its shareholders and customers.
In the next few years, we’ll have a new generation of IT leadership that’s better equipped to articulate that value. That leadership will come from individuals like today’s chief technology officers who are evaluating and experimenting with new technologies like containers and IoT, and the new wave of chief product officers who are starting to emerge since we live in an age when a business’s products themselves are software in some shape or form.