2021 in Cloud Native: Kubernetes Goes Invisible, Ops Goes Distributed, and Infrastructure Gets Easier

By Bruno Andrade, CEO of Shipa.io

I think 2021 will be a particularly transformative time for enterprise DevOps teams, as prominent infrastructural technologies fade into the background, and as organizations look for ways to restructure and redefine Ops teams and practices for better workflow efficiency.

More specifically, I anticipate that:

1) For DevOps teams, Kubernetes will shift from an inescapable focus to an afterthought.

Kubernetes - now the ubiquitous mainstay of modern software deployment - is surprisingly primed to perform a disappearing act. Kubernetes offers countless positives in standardizing how DevOps teams approach the packaging, running, and monitoring of workloads. But it also has limiting pain points, including particularly challenging complexity and pervasive operational inefficiencies. The opaqueness of Kubernetes makes maintenance and troubleshooting issues that much tougher to resolve, and provides dark corners where security threats may lurk until they're ready to strike.

Ironically, that lack of visibility will contribute to Kubernetes ultimately becoming a transparent commodity - one that's out-of-sight of DevOps teams. Operating Kubernetes internally will simply become too time-consuming, resource-intensive, and burdensome to justify. Instead, enterprises will increasingly reap the advantages of relying on Kubernetes managed services and automated frameworks in 2021, taking the opportunity to reallocate their own resources to product growth and other business initiatives.

2) Enterprises will find a new home for Ops: inside development teams.

Ops has been long due for a bold reimagining. Developers' requests will always - always - exceed Ops capacity. Enterprises also face a fundamental challenge in that Ops cannot be scaled without endlessly adding expensive (and often hard to find) personnel. Put shorter, Ops cannot be scaled. As a result of these factors, organizations suffer from "Ops lock-in," with Ops impeding innovation by being unable to keep pace with development demands. At the same time, Ops teams are responsible for everything, but often don't have the specific on-the-ground knowledge to fix issues directly. Because of this, Ops spends significant time identifying and routing issues to personnel that do have the correct knowledge. Those individuals are often developers.

In 2021, expect organizations to address these limitations by embracing distributed Ops, which places Ops duties with the development teams most capable of solving Ops issues specific to their projects. The future of Ops will give developers responsibility over their own services, and powerful end-to-end self-service tools and options for managing them effectively.

3) Enterprises will simplify developers' interactions with infrastructure (especially by embracing DevOps process workflows).

Developers are now required to deal with infrastructure increasingly often. However, those interactions take inexperienced (and even many experienced) developers out of their comfort zone, resulting in frustration, distraction, less productivity, and more mistakes. Organizations will address these pain points throughout 2021 by pursuing strategies that can simplify and abstract away infrastructure's challenging aspects.

Today, (most) developers lack true workflows for the DevOps process - workflows that, if executed properly, would profoundly accelerate their progress. Improving this inefficiency in 2021, expect developers to be able to start zeroing in on application code more, rather than dealing with relative minutiae like creating objects. Rules defined at the framework level will be automatically enforced, and platform teams can then work on controls rather than building and maintaining custom scripts. Circling back to Kubernetes as an example, expect organizations to introduce intermediate layers between the orchestrator and applications, such that developers receive the full benefits of Kubernetes while being allowed to focus simply on pushing code. Organizations will be able to draw a bright line dividing the work of DevOps and platform engineering teams managing infrastructure, and developers rapidly iterating applications upon that infrastructure while taking it for granted.

##

About the Author

Bruno Andrade is the CEO of Shipa.io, which makes a cloud native application management framework to manage the full application lifecycle. Using Shipa, organizations speed up cloud native application development by eliminating persistent workflow inefficiencies. Previously, Bruno held engineering leadership roles at Juniper Networks, Oracle, and IBM, and was the CEO of HTBASE. He lives in Mountain View, California.