How to Use Metrics Server in Kubernetes for Resource Monitoring

To use metrics-server in Kubernetes, first deploy it to your cluster to collect resource usage data. Then use commands like kubectl top nodes or kubectl top pods to view CPU and memory metrics in real time.

Syntax

The metrics-server is deployed as a Kubernetes add-on that collects resource metrics from nodes and pods. After deployment, you use kubectl top commands to query metrics.

• kubectl top nodes: Shows CPU and memory usage for each node.
• kubectl top pods: Shows CPU and memory usage for pods in the current namespace.
• kubectl top pods --all-namespaces: Shows metrics for pods across all namespaces.

bash

kubectl top nodes
kubectl top pods
kubectl top pods --all-namespaces

Example

This example shows how to deploy the Metrics Server and then check resource usage of nodes and pods.

bash

kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

# Wait a few moments for metrics-server to start
kubectl top nodes
kubectl top pods --all-namespaces

Output

NAME           CPU(cores)   MEMORY(bytes)
node-1         250m         512Mi
node-2         300m         600Mi

NAMESPACE     NAME                             CPU(cores)   MEMORY(bytes)
default       my-app-pod-1                     100m         200Mi
kube-system   metrics-server-abcdef12345       5m           20Mi

Common Pitfalls

Common mistakes when using Metrics Server include:

• Not deploying Metrics Server before running kubectl top commands, which results in errors like "metrics API not available".
• Metrics Server failing to start due to missing permissions or network issues.
• Using an outdated Metrics Server manifest that is incompatible with your Kubernetes version.

Always use the latest official Metrics Server manifest and check pod logs if metrics are missing.

bash

kubectl top nodes
# Error: metrics API not available

# Correct approach:
kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml
kubectl top nodes

Quick Reference

Command	Description
kubectl apply -f	Deploy Metrics Server to the cluster
kubectl top nodes	Show CPU and memory usage of nodes
kubectl top pods	Show CPU and memory usage of pods in current namespace
kubectl top pods --all-namespaces	Show pod metrics across all namespaces
kubectl get deployment metrics-server -n kube-system	Check Metrics Server deployment status

Key Takeaways

Deploy Metrics Server using the official manifest before querying metrics.

Use 'kubectl top nodes' and 'kubectl top pods' to view resource usage.

Ensure Metrics Server pods are running and have proper permissions.

Use the latest Metrics Server version compatible with your Kubernetes cluster.

Check Metrics Server logs if metrics are missing or commands fail.