August 13, 2026
K8s, chaos, and hardware thirst
Kubernetes on Oxide: How Customer Needs Shaped Our Integrations
Customers begged for an easier way to run apps on Oxide, and the comments got thirsty fast
TLDR: Oxide says customer demand pushed it to finally build easier ways to run Kubernetes on its systems, starting with a customer-submitted fix and expanding into multiple integrations. The community loves the customer-first approach, but the comments mix hype, jokes, and a clear warning that Oxide still has catching up to do.
Oxide’s latest update is basically a backstage story about customers dragging the roadmap into existence. The company says people desperately wanted to run Kubernetes — a popular system for managing lots of apps across machines — on Oxide hardware, but the tools just weren’t there yet. So what kicked this whole thing off? Not some grand top-down master plan, but a customer pull request and a rough internal draft. From there, Oxide built several ways to get clusters up and running, following real user pain points instead of guessing.
And honestly, the comment section is having a blast with that. One crowd is fully in “let them cook” mode, with people praising Oxide’s engineering style and already joking about future add-ons like a hypothetical “karpenter-provider-oxide” showing up on someone’s bingo card. Another group is loudly thirsting over the company’s whole vibe: one commenter said they’d “absolutely kill” for Oxide to open-source its documentation system, while another went full sci-fi yearning and confessed they want an Oxide rack at home, maybe after spotting one in a surplus auction in 40 years. That’s not feedback — that’s fandom.
But there’s also a little edge. Not everyone thinks Oxide is fully caught up yet. One commenter argued this is exactly where Oxide still doesn’t quite match the public cloud, saying the setup still feels less seamless than what big cloud providers offer. So the mood is equal parts admiration, wishlist, and gentle side-eye: Oxide is winning fans by listening, but the audience is absolutely keeping score.
Key Points
- •In late 2024, Oxide had customer demand for Kubernetes on its platform but no supported integrations.
- •The company’s Kubernetes integration work began with a customer-submitted Rancher node driver pull request and an early draft of RFD 493.
- •Oxide merged, improved, documented, and released the Rancher node driver, which became its first official Kubernetes integration and was already in production use by a customer.
- •Oxide determined that one provisioning workflow would not fit all customers and ultimately published three integrations.
- •Oxide later partnered with Sidero Labs to build an Oxide infrastructure provider for Omni, uncovering compatibility issues in Omni and Talos Linux during the effort.