k8s: How the controller-runtime Cache Actually Works, and Why Your Controller Does Not Crash the API Server

45 · Sujith Quintelier · July 30, 2026, 7:42 a.m.
Summary
This blog post explains how controller-runtime controllers in Kubernetes utilize a local informer cache to manage reads and writes. The author details the mechanisms behind shared informers, DeltaFIFO, indexing, and cache scoping, which impact consistency and memory usage. The piece offers insights into caching behavior and API interactions that can enhance understanding of Kubernetes architecture.