This year, CircleCI CTO Rob Zuber predicts DevOps professionals will take a more incremental approach to workflow acceleration.
“We will see teams getting really practical and focusing on specific tactics like continuous integration and continuous delivery,” said Zuber.
This laser-like focus on specific tactics will help developers be more mindful of potential risks when making decisions, Zuber said, and ultimately help deliver values to users and the business faster. Additionally, Zuber predicts that with the increased adoption of microservices, the practice of testing in production will increase as well.
Reshaping thinking around production validation doesn't mean being dismissive or reckless, however.
“It means shifting engineering focus and resources to limiting blast radius and speeding up production bugs remediation,” Zuber said.
What are the top tech trends you’re watching in 2020?
Incremental advancements in DevOps will prove more successful than the all-or-nothing approach. We will see teams getting really practical and focusing on specific tactics like continuous integration and continuous delivery.
Developers will be more attuned to risk in their decision-making. They will use tools and resources to help them deliver value to users faster, find product-market fit sooner and stay focused on creating business value.
Testing in production will increase. With the increased adoption of microservices and distributed systems, the cost of catching specific classes of issues in a pre-production environment has grown. It stands to reason that these costs can and will theoretically increase to the point where the lower-cost, lower-risk option is finding issues in production instead and fixing them as quickly as possible.
This is where testing in production outgrows its history as a meme and takes its place as a viable methodology. However, reshaping our thinking around production validation doesn’t mean being cavalier. It means shifting engineering focus and resources to limiting blast radius and speeding up production bugs remediation.
Teams will expand their use of tools that give them production confidence through better exception handling and observability. And when they identify production issues, they will be even more dependent on rapid resolution through solid CI/CD pipelines.
This is where testing in production outgrows its history as a meme and takes its place as a viable methodology.’’
How are you applying these trends in your work in the year ahead?
As a software development community, to reduce the risk of any one change, we have embraced the process of slicing changes into the smallest possible increments. The tooling that allows us to do that, such as CI/CD with comprehensive test coverage, has given us the confidence to move quickly. We know we will not deploy anything to a production environment until it is known to be good.
As software becomes a competitive differentiator for more and more companies (not just in tech but in retail, health and finance), software teams are optimizing for faster and faster delivery. Trends that have appeared alongside these goals include greater use of third-party services and tools, microservice architectures and larger data sets.
As we’ve learned to get better and better at risk mitigation in software delivery, our failures get smaller and more comfortable. This year, I’ll not only be thinking about how we can embrace risks in software delivery, but also how my team can embrace risks in their own personal development.
