DevOps Glossary

Kubernetes Pod Topology Spread Constraints are scheduler rules in Kubernetes that tell the control plane how evenly to place matching Pods across topology domains, such as nodes, zones, regions, or racks. They help prevent too many replicas of the same workload from landing in one failure domain by using settings such as maxSkew, topologyKey, whenUnsatisfiable, and a label selector. For example, a web Deployment with 6 replicas can use a zone topology key to keep Pods balanced across 3 availability zones, reducing the chance that one zone outage takes down most of the service. They are useful for highly available services, multi-zone clusters, and large clusters where default scheduling may create uneven placement, but they can leave Pods pending if the cluster lacks enough capacity in the required domains.