Kubernetes is Your Cloud Control Plane

Kubernetes is a cloud control plane that provides a standard interface for managing compute, network, and storage in any cloud or datacenter. It aims to deliver high-level capabilities on any underlying provider via the same consistent interfaces used for compute, network, and storage.

• Kubernetes API and custom resource definition emerged as an open standard for describing cloud infrastructure, services, and apps
• Paradigms and frameworks used to build and manage controllers for these resources
• Conventions emerging from projects like Crossplane and Operator Framework to provide consistency and simplicity for developers and operators of custom resources
• Common problems that Kubernetes resource providers must handle, such as publishing connection secrets and managing controller provisioning and updates

The speaker shared his experience working in open standards and open source for over a decade, including his work on Swagger spec, OpenAPI, and DevOps specs. He also worked in Azure for a long time and pushed for their resource manager spec to become an open standard. He now works for Red Hat, which uses Kubernetes as the control plane for everything.

Kubernetes accelerates app development by providing a standard interface for managing compute, network and storage in any cloud or datacenter. But the promised value of cloud computing depends on more than these; developers need higher-level services too, like databases, streams, buckets, identities, monitors and pipelines. And so Kubernetes' next act is to deliver high-level capabilities on any underlying provider via the same consistent interfaces used for compute, network and storage. That is, soon every dependency of your app will be described and managed by Kubernetes resources and their controllers.In this session we'll describe how the Kubernetes API and custom resource definition emerged as an open standard for describing cloud infrastructure, services and apps. We'll discuss paradigms and frameworks used to build and manage controllers for these resources and compare and contrast types published by several providers, like Strimzi's KafkaTopic, cert-manager's Certificate or a DynamoDB Table. We'll demonstrate conventions emerging from projects like Crossplane and Operator Framework to provide consistency and simplicity for developers and operators of custom resources. And we'll review common problems that Kubernetes resource providers must handle, such as publishing connection secrets and managing controller provisioning and updates.