LONDON — With Kubernetes and the cloud native community now 10 years old, KubeCon was a chance to both look back and look ahead for Google’s Jago Macleod, engineering director of GKE and Kubernetes at the company.
In a fascinating presentation at KubeCon + CloudNativeCon Europe in London this week, Macleod explained Google’s initial motivation for open sourcing Kubernetes, how it evolved over the following decade and — most interestingly — how it plans to accommodate the changes wrought by generative AI.
Here are the highlights in 10 slides, with quotes from Macleod to accompany each image.
“The first phase was really about disruption,” said Macleod, about the initial phase from 2014–2017. “This was us [Google] entering the public cloud market.”
He added a bit later: “We felt we had an opportunity to expand the public concept of cloud from just VMs to also containers.”
The era of 2018 to 2022 “was the ecosystem expansion.”
“And there was an explosion of ecosystem projects and solutions to problems,” Macleod continued. “Istio comes from that time, the OPA Gatekeeper, stuff like Argo and Knative, and OpenTelemetry. And so the more users there are, the more valuable it is to create for this ecosystem.”
The next phase, Macleod thought, was “going to be about consolidation.”
“With all those ecosystem projects, we have to provide opinions — not just options. We got feedback from end users that it was just too confusing, too complex. And I thought it was going to be about stability, simplicity, being comprehensive. You need a whole platform. You don’t just need one part from a bag of Legos. You want the toy. And so that’s what I thought the future was going to be like.”
“And then there was this plot twist at the end of 2023, right? The ChatGPT moment. And the whole world overnight was just obsessed with this idea.”
Macleod noted that not only was AI being a major disrupter, but that there was still natural increasing demand for Kubernetes on the enterprise side.
“Demand outstripped supply quite significantly,” he said of this period, from 2023 on. “And so we have these crazy overlapping stacking adoption curves. And it’s really interesting, there’s […] enterprises still coming to cloud native to learn about containerization and microservices, and then there’s an entirely different plane [AI] that’s trying to upend the whole thing at the same time.”
In response to the double growth curve, Macleod continued, “we came up with these three stories to rally around inside of Google.” The goal was “to evolve Kubernetes to meet the needs of the next trillion core hours.” The three goals are:
- Improve reliability at scale, across upgrades.
- Redefine Kubernetes’ relationship with hardware.
- Move from container and workloads to framework orchestration.
“The idea of Kubernetes has been based on the hourglass models,” Macleod then explained. “And this was a diagram that circulated, I don’t know, 40 years ago. The whole concept is that, like, IP is the center — the narrow waist — driving the ecosystem on top and different technologies. […] And so making Kubernetes the narrow waist of the hourglass model of infrastructure consumption is our vision.”
From the hourglass model, Google derived its ongoing strategy for Kubernetes.
“Number one,” said Macleod, “maintain to ensure that Kubernetes continues to thrive. Extend Kubernetes as the de facto standard for infrastructure, and expand it — especially for AI/ML workloads that are so important to every business. Make sure that it works for those frameworks and the new workloads that are coming as well.”
The overall goal is that Kubernetes becomes “GenAI aware,” as one of Macleod’s slides put it.
Of course, as a business operating on top of Kubernetes, Google wants to provide points of differentiation. Essentially, Macleod explained, it’s all about Google’s ability to scale its public cloud.
“If we’re doing all this in open source, to be totally open and honest, our intention is to differentiate on performance, really — and then that gives us business opportunities to do price performance [options]. You can offer the same performance [as open source K8s] for a lower price, or better performance [for a higher price] — like, there are options, but really it’s about differentiation on performance.”
Naturally, Macleod is optimistic about the future of Kubernetes, given the current upswing in demand from both enterprise adoption maturity and the generative AI boom.
“Kubernetes is declarative, extensible and modular,” he said. “That makes it super well positioned to evolve to meet the needs of the next round of workloads.”
Macleod concluded by noting that Kubernetes is poised for further expansion.
“There’s a lot invested in it. There’s a lot of value. And so we have a window of opportunity. We really do believe this is a window of opportunity to evolve so that we don’t get pushed out of the way in favor of something else. We’re well on our way. We’re accelerating, we’re speeding up. We’re actually investing a lot more in Kubernetes than we were a year or two ago. We think you should too, and stoked to see what we can build together.”
Originally published at The New Stack: https://thenewstack.io/kubecon-europe-how-google-will-evolve-kubernetes-in-ai-era/