Platform and infrastructure teams often rely on ad-hoc scripts to manage operational knowledge—until complexity, scale, and reliability demands break those workflows. In this session, the speakers will present how to replace fragile operational scripts with a purpose-built Kubernetes controller to handle real-world workflows as native Kubernetes resources.
This session will include a walkthrough of designing your own controller using controller-runtime, writing reconciliation loops, and ensuring safe, idempotent behavior in production. The speakers will share lessons from building a controller that coordinates multi-step workflows (such as upgrades, or compliance checks), and handles failure gracefully, with appropriate re-queueing and controller concurrency.
Attendees will walk away with a clear design patten for deciding when a Kubernetes controller is the right choice—and when it’s not—plus actionable patterns for replacing brittle scripts with resilient, reusable automation.