Monitor Kubernetes Resources with kubectl top 

Monitoring Kubernetes resources is a critical task for maintaining cluster performance and ensuring efficient resource usage. Kubernetes provides a native tool, kubectl top, which allows you to monitor resource usage (CPU and memory) of nodes and pods in real-time. This guide explains how to use kubectl top effectively.

Step 1: Prerequisites

Before using the kubectl top command, ensure the following:

1. Metrics Server:

Kubernetes requires the Metrics Server to collect and provide resource metrics. Check if Metrics Server is installed:

kubectl get deployment -n kube-system

Look for metrics-server in the output. If not installed, you can set it up by following the Metrics Server installation guide.

2. RBAC Permissions

Ensure your user or role has the necessary permissions to access metrics. You can define a ClusterRole as follows:

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: view-metrics
rules:
  - apiGroups: [""]
    resources: ["pods", "nodes"]
    verbs: ["get", "list", "watch"]

Step 2: Monitor Node Metrics

To view CPU and memory usage across all nodes in your cluster, run:

kubectl top nodes

Example Output:

NAME             CPU(cores)   CPU%   MEMORY(bytes)   MEMORY%
node1            250m         5%     1Gi             25%
node2            300m         6%     1.5Gi           37%

This provides a quick overview of the cluster’s resource utilization.

Step 3: Monitor Pod Metrics

For a Specific Namespace: To see resource usage for pods within a namespace:

kubectl top pods -n <namespace>

For All Namespaces: To list resource usage across all namespaces:

kubectl top pods --all-namespaces

Example Output:

NAME              CPU(cores)   MEMORY(bytes)
nginx-pod         100m         200Mi
api-server-pod    500m         600Mi

Step 4: Monitor Specific Resources

Monitor a Specific Pod:

kubectl top pod <pod-name> -n <namespace>

Monitor a Specific Node:

kubectl top node <node-name>

ip-10-0-4-216.ap-south-1.compute.internal   574m         29%    4099Mi          57%       
ip-10-0-5-109.ap-south-1.compute.internal   415m         21%    2978Mi          41%       
ip-10-0-6-156.ap-south-1.compute.internal   615m         31%    6028Mi          84% 

Step 5: Automate Monitoring with Continuous Updates

To monitor resource usage continuously, combine kubectl top with the watch command:

watch -n 5 kubectl top pods -n <namespace>

NAME                                            CPU(cores)   MEMORY(bytes)
aws-load-balancer-controller-76f7d9b8f7-c2bfr   4m           32Mi
aws-node-2bctx                                  4m           40Mi
aws-node-6hhpb                                  5m           35Mi
aws-node-ptdhg                                  6m           36Mi

This updates the output every 5 seconds.

Why Use kubectl top?

• Real-Time Monitoring: Provides immediate insights into resource usage.
• Native Integration: No need for additional tools to start monitoring.
• Quick Debugging: Ideal for troubleshooting resource contention issues.

For advanced monitoring, consider integrating tools like Prometheus and Grafana for detailed metrics visualization and alerting capabilities.

Conclusion

Using kubectl top is a simple yet powerful way to monitor Kubernetes resources. It’s especially useful for quick checks and troubleshooting. For more comprehensive monitoring and visualization, pairing it with tools like Prometheus and Grafana is recommended. Start using kubectl top today to keep your Kubernetes cluster running smoothly.