Kubernetes Pod Topology Spread Constraints
Kubernetes Pod Topology Spread Constraints are scheduler rules that spread the pods of one workload evenly across groups of machines such as nodes or zones. You choose the grouping with a topology key and set maxSkew, the largest allowed difference in pod count between groups. For example, 6 replicas across 3 zones land as 2 per zone, so the loss of 1 zone removes only 2 pods.

See also Kubernetes PodDisruptionBudget (PDB), Kubernetes PriorityClass, Kubernetes Readiness Probe, Kubernetes ReplicaSet, Kubernetes Resource Requests and Limits, Kubernetes ResourceQuota.