Chapter 15: Kubernetes Introduction
15.12c Resource requests (awareness)
Production clusters set CPU/memory requests/limits so the scheduler places Pods fairly. On a tiny kind lab you may omit them at first — but know the words for interviews: without requests, noisy neighbours can starve others; without sensible limits, one leak can hurt the node.
Do not invent cluster-size capacity numbers here. On kind, watch Docker Desktop/engine disk and memory; delete clusters when done.