LimitRanges and ResourceQuotas

One careless Deployment with no resource limits and the cluster has a noisy neighbor. Without namespace-level controls, Kubernetes schedules pods wherever they fit and lets them consume whatever node capacity they can reach. Teams sharing that cluster don't find out until pods are evicted or latencies spike at 2 AM.

LimitRanges and ResourceQuotas are the namespace-scoped policy objects that prevent this. The difference is scope: LimitRange is per-object, ResourceQuota is per-namespace. They're enforced sequentially at admission on every pod CREATE — LimitRange mutates first, then both validate.