Kubernetes v1.36: Pod-Level Resource Managers (Alpha)
Kubernetes v1.36 adds Pod-Level Resource Managers as an alpha feature, extending the kubelet's Topology, CPU, and Memory Managers to use pod-level resource specifications. Exclusive NUMA-aligned resources can be granted to specific containers while sidecars share a pod-level pool, under feature gates PodLevelResourceManagers and PodLevelResources. New kubelet metrics track allocations, errors, and container assignment types.
What changes for platform teams
Platform teams running latency-sensitive workloads can keep sidecars in the Guaranteed QoS class without dedicating exclusive cores to them, but must enable two feature gates and configure the static CPU and Memory Managers with a non-none Topology Manager policy. Since the feature is alpha, teams should review the documented limitations and downgrade instructions before relying on it.
Read the original at Kubernetes Blog ↗
written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules