808labs

Scoring the 2015 Kubernetes Predictions: What a Community Can Build

—

by

Barton George
in

With KubeCon North America just days away, I decided to revisit the promises, predictions and goals laid out when Kubernetes 1.0 and the CNCF were announced at OSCON 2015. Which were delivered on, which fell short, and which were abandoned? To find out, I went back to my blog post from that day, which includes an interview with Kubernetes co-founder Craig McLuckie and links to three contemporary articles.

The scorecard

Here’s my unofficial scorecard, drawn from that interview and the articles published at the time. Spoiler alert: they’ve done pretty well.

That’s impressive given that no one has ever accused Kubernetes of being easy to understand or work with. And yet it has spawned an entire ecosystem: the first KubeCon, in 2015, drew about 500 people, and today the big KubeCons draw 12,000 to 13,000.

Support for new workloads

2015 goal: Add support for big data, batch processing and high-throughput computing.

Result: Accomplished.

  • Batch jobs, Spark and AI training all run on Kubernetes today.
  • CNCF now calls it the de facto operating system for AI.

Scaling up

2015 goal: Go from a few hundred nodes per cluster to thousands.

Result: Accomplished.

  • Upstream Kubernetes supports 5,000 nodes.
  • Amazon EKS supports up to 100,000 nodes per cluster.

Community growth

2015 goal: Increase both commits and contributors.

Result: Accomplished.

  • 2015: 14K commits, 400+ contributors.
  • By 2023: 314K+ commits, 74K+ contributors.

Broad adoption of cloud native

2015 goal: Support adoption of cloud native computing technologies.

Result: Accomplished.

  • 82% of container users now run Kubernetes in production (CNCF survey, Jan 2026).
  • Kubernetes has become the standard underlying platform for running modern applications.

Expansion beyond Kubernetes

2015 goal: Include new projects, tools, APIs and standards for containers and cloud infrastructure.

Result: Accomplished.

  • From 1 project to 230+, including Prometheus, Envoy, Helm, Argo, OpenTelemetry and Kubeflow.

Holdouts joining

2015 prediction: Microsoft, Amazon, Pivotal and other holdouts will join later.

Result: Accomplished.

  • Microsoft, AWS and Pivotal all joined in 2017.
  • Membership went from 22 founding members to nearly 800.

An open source case study

Beyond hitting its goals, Kubernetes also shows what open source and the broader community can deliver. A few of the benefits it illustrates:

  • Collaboration across competitors. Companies with competing goals, many of them direct rivals, worked together to build out a shared platform, then competed on what they built on top of it.
  • Adding the features that matter to you. When Kubernetes launched, it was missing a fair amount of functionality, particularly as it related to the enterprise. Companies worked with the community to add those capabilities, increasing the platform’s usefulness for everyone. Storage vendors, for example, helped develop the Container Storage Interface (CSI), extending Kubernetes to support enterprise storage systems customers already owned. Vendors gained access to a growing market, while Kubernetes became useful to more enterprises.
  • Standards driven by adoption. Rather than waiting for a traditional standards body to define how containers should be managed, the Kubernetes community built working software and let adoption drive convergence. Kubernetes became the common platform across the major clouds without a formal standards body selecting it.
  • Reshaping markets. By open sourcing Kubernetes and building a community around it, Google helped move the industry toward cloud-native architecture and away from the VM-centric approach that had defined enterprise infrastructure.

My favorite prediction

To return to the scorecard, the last item, the holdouts, is my favorite. At the 2015 launch, TechCrunch noted that Microsoft, Amazon and Pivotal were missing from the new foundation, but the Linux Foundation’s Jim Zemlin wasn’t worried. “I predict this technology will be too good to resist,” he said. If that doesn’t sound like typical executive director puffery, I don’t know what does. This time, though, he was right: it was too good to resist. By 2017, all three had joined.

From technology to platform

Great technology alone doesn’t create a successful platform. It needs to be useful and widely adopted. This is where open source can play an important role. When a community forms around the technology, its members can expand its capabilities and utility. This, along with the assurance of ongoing support, helps drive adoption. Kubernetes, with the support of the CNCF, shows how a community with shared interests can take an initial contribution, build on it and turn it into a platform that helps shape the direction of an industry.

Pau for now…

Sources

Comments

Leave a Reply