As Kubernetes evolves beyond 5yrs of adoption, many say that it is becoming “boring”. So where does the popular open source project and community go next?
SHOW: 509
SHOW SPONSORS:
Join Turbonomic, Kelsey Hightower, and Corey Quinn for the Apps ON Cloud Summit May 11-13. Swag & giveaways!
If Kubernetes is boring, what’s next for Kubernetes?
Kubernetes has now shipped 21 releases, which is nearly as many as OpenStack (24 releases), and more than VMware vSphere (12-15 releases).
Kubernetes is shifting from 4 releases a year to 3 releases a year.
On average, 170-180 companies are contributing monthly.
It’s widely available both in private cloud and public clouds
Large percentage of larger companies have projects running on Kubernetes
So what might be next?
To some extent, the projects surrounding and adjacent to Kubernetes are gaining attention and momentum - Serverless, Service Mesh, Multi-Cluster Management Tools, Security Tools, Operators, Helm Charts
Kubernetes is being used across Private Clouds, Public Clouds, a variety of Edge use-cases, and is gaining adoption by Telcos for 5G/Edge use-cases.
Beginning to see some thinking and frameworks around how to use Kubernetes as a Hybrid/Multi-Cloud control plane, not just for Cluster deployment.
Edge use-cases are interesting, since Edge has so many varied deployment models - Does Kubernetes need to be miniaturized? Does Kubernetes ship too often?
AI/ML/Analytics use-cases are exploding as most frameworks are making Kubernetes the default orchestration framework.
Gitops is starting to gain some momentum as an automated, single-source-of-truth management model for Kubernetes clusters and application deployment.
The new way of companies using the public cloud are now trying to figure out how to manage Cloud costs, especially as Kubernetes drives more self-service engagement with platforms.