dcsimg
Read more on "Real World Open Source" »

Why Kubernetes Is Successful and Boring

By Sean Michael Kerner (Send Email)
Posted December 12, 2018


SEATTLE — Google has had a common message throughout 2018 about Kubernetes, and the message is simple: Kubernetes is boring.

At the KubeCon + CloudNativeCon NA 2018 event here, Google engineer and conference co-chair Janet Kuo echoed comments made by her peer Aparna Sinha, group product manager at Google, at the Kubecon and CloudNativecon Europe 2018 keynotes in May, which is simply that Kubernetes is boring, and boring is good.

Kuo said in the early days of Kubernetes the focus was on building fast and adding new features. By 2015, a focus was added to make it easier for users and administrators to build, deploy and use Kubernetes. At this point in the maturity cycle of Kubernetes, Kuo commented that adoption has moved from the early stage of adopters to more mainstream deployments.

"Kubernetes is now getting so solid and so mature and so great, that it is very, very boring," Kuo said during her keynote. "Boring is good; it means that lots of companies are already using it, and it just works."

Kuo added that being boring means organizations can just focus on delivering business value, rather than spending time on making Kubernetes usable.

Standardized Built-in APIs for Kubernetes

There are a number of things that make Kubernetes stable and usable at this point. At the core of Kubernetes is a set of standardized built-in APIs that provides a layer over infrastructure. That abstraction layer enables Kubernetes to run in all sorts of different environments, ranging from bare metal, to public, private and edge cloud deployments.

Standards for Kubernetes is also about conformance. The Cloud Native Computing Foundation (CNCF) has developed a Kubernetes conformance testing program to make sure products and services that say they are Kubernetes, actually are Kubernetes and work in a standardized way.

"We want your workload to work anywhere and can be deployed correctly in a given implementation," Kuo said. "With conformance, you don't have to test against every Kubernetes environment, so when you see the Kubernetes conformance logo, you know you're getting consistent behaviors."

Kubernetes' Extensibility for Both Infrastructure and API Layers

Another key attribute of Kubernetes that has made it successful is the fact that it is extensible for both infrastructure and API layers.

Kuo explained how infrastructure extensibility enables organizations to customize the way Kubernetes consumes the underlying infrastructure. As such, Kubernetes can be configured to leverage different cloud providers, all via standardized plugin APIs.

Interfaces such as the Container Networking Interface (CNI) and Container Storage Interface (CSI) eliminate barriers to entry for new service providers and help to build a healthier ecosystem, according to Kuo.

Kubernetes also provides API extensibility. Kuo said that even though the core Kubernetes API covers 80 percent of use cases, the API extensibility capability empowers organizations to create their own custom APIs and controllers to help automate things.

"Kubernetes' open standards and extensibility make it a ubiquitous platform for building your cloud-native platform and provide a standardized abstraction, with consistent behavior because of conformance," Kuo said. "Boring is necessary because you need it to be boring in order to have a mature platform to build and depend upon."

Sean Michael Kerner is a senior editor at ServerWatch and InternetNews.com. Follow him on Twitter @TechJournalist.

Page 1 of 1

Read more on "Real World Open Source" »
Thanks for your registration, follow us on our social networks to keep up-to-date