Back to Glossary
Core Concepts

Control Plane

The set of components that store cluster state, schedule workloads, and run controllers to maintain desired state.

What is Control Plane?

The Kubernetes control plane is the brain of a cluster. It consists of four core components: the API server (kube-apiserver), which is the single entry point for all API calls and enforces authentication, authorization, and admission control; etcd, a distributed key-value store that holds all cluster state; the scheduler (kube-scheduler), which assigns Pods to nodes based on resource availability, affinity rules, and policies; and the controller manager (kube-controller-manager), which runs reconciliation loops for objects like Deployments, ReplicaSets, and Nodes.

On managed Kubernetes services (GKE, EKS, AKS), the control plane runs on cloud-provider infrastructure and is billed separately — typically $72–$146/month per cluster for GKE and EKS. The API server is highly available via leader election, and etcd is replicated across at least three nodes in production setups. All cluster state changes — creating a Pod, scaling a Deployment, updating a ConfigMap — flow through the API server and are persisted in etcd.

etcd performance degrades with large object counts or large individual objects. Clusters storing thousands of unused Secrets, ConfigMaps, or large CRD instances in etcd experience slower API response times, which degrades kubectl operations and controller reconciliation speed for all users of the cluster.

Example

# Check control plane component statuses
kubectl get componentstatuses

# View kube-system pods (control plane components)
kubectl get pods -n kube-system

# Check API server audit log (if enabled)
kubectl logs -n kube-system kube-apiserver-my-node

Cost & Waste Implications

Managed control planes are a fixed cost per cluster — you pay whether the cluster runs 1 Pod or 10,000. Accumulating thousands of unused objects (orphaned ConfigMaps, completed Jobs, stale Pods) bloats etcd and degrades API server performance for everyone, effectively increasing the operational cost of running the cluster without adding capacity.

KorPro— Kubernetes Cost Optimization

How KorPro Helps

KorPro reads from the Kubernetes API (read-only) without adding load to your control plane, and highlights etcd-bloating patterns like large numbers of completed Jobs, stale ConfigMaps, and unused Secrets.

Scan Your Cluster Free

Stop Wasting Money on Orphaned Kubernetes Resources

KorPro connects to your clusters across GCP, AWS, and Azure — no agents, no installation — and surfaces every orphaned resource with its monthly cost estimate.