KubeVirt Graduates After a Decade of Uniting VMs and Containers
11m
KubeVirt has reached CNCF graduation, ten years after its first commits. The project set out to run virtual machines on Kubernetes. Today it sits alongside Kubernetes and Prometheus at the foundation's highest maturity level.
KubeVirt Founder and Maintainer Fabian Deutsch examines that milestone with Alan Shimel at KubeCon + CloudNativeCon in Salt Lake City.
A Bet on Kubernetes
Around 2016, the winner of the orchestration race was far from obvious. Docker Swarm, Mesosphere and Rancher were all in the mix. Red Hat made an educated bet on Kubernetes, built on open source trust and a shared vision with Google.
That raised a hard question for virtualization teams. Did enterprises really want to maintain two platforms? One would run containers, and another would run virtual machines.
One Layer Instead of Two
VMs are not going away. Cost, legal requirements and existing applications all keep them in service. The goal was to make running them simpler for the whole company, not just one administrator.
Deutsch flips the classic picture. Instead of Kubernetes running inside a virtualization stack, the container platform sits at the bottom. VMs blend into it when needed.
Stacking platforms means solving networking, storage, observability and identity more than once. It also means training people twice. Teams that already understand Kubernetes face only a small learning delta.
AI, Security Reports and Graduation
AI has brought a surge of contributions and vulnerability reports to open source projects. Each finding still needs a person to confirm it is real. At the same time, AI tools are helping maintainers support their feature process.
Graduation required a third-party security audit, due diligence and a review by the CNCF Technical Oversight Committee. Deutsch says it reassures end users and reflects a community that now carries the load on many shoulders.
Newcomers can start at KubeVirt.io with the getting started guide, the weekly community call and the Slack channel. Documentation, AI policy work and code contributions are all welcome.