Kubernetes Node Provisioning & Pod Distribution

Monitor Kubernetes node provisioning and pod distribution across EKS, AKS, and GKE with support for both managed nodegroups and Karpenter-provisioned nodes on any cloud.

Metrics displayed:

Total nodes provisioned per nodegroup or node pool Total pods running on selected nodes Total DaemonSets in the cluster Pods scheduled by DaemonSets per node Kube-system pods per node Pod-level CPU and memory usage, requests, and limits

Filter by:

Node Group (EKS managed, Karpenter, AKS agent pool, GKE node pool) Node Namespace Workload Type (Deployment, StatefulSet, DaemonSet, etc.)

Requires kube-state-metrics with node labels exposed:

kube-state-metrics: 
  metricLabelsAllowlist:   
  - nodes=[*]

Also requires the namespace_workload_pod:kube_pod_owner:relabel recording rule, which is included by default in kube-prometheus-stack.