Resource monitoring best practices in Kubernetes - Commands & Configuration

Introduction

Monitoring resources in Kubernetes helps you keep your applications healthy and running smoothly. It shows you how much CPU and memory your apps use, so you can fix problems before they get big.

When you want to know if your app is using too much CPU or memory and might slow down.

When you need to find out why your app is crashing or restarting often.

When you want to plan for more servers because your app is growing.

When you want to see if your app is balanced well across your servers.

When you want to get alerts if your app's resource use is too high.

Commands

This command shows the current CPU and memory usage of all pods in the default namespace. It helps you see which pods use the most resources.

Terminal

kubectl top pods

Expected Output

NAME           CPU(cores)   MEMORY(bytes)

my-app-12345   50m          100Mi

my-db-67890    30m          200Mi

This command shows the CPU and memory usage of each node in your cluster. It helps you understand the overall resource use on your servers.

Terminal

kubectl top nodes

Expected Output

NAME           CPU(cores)   MEMORY(bytes)

node-1         500m         2Gi

node-2         300m         1.5Gi

This command gives detailed information about a specific pod, including its resource requests and limits. It helps you check if the pod has proper resource settings.

Terminal

kubectl describe pod my-app-12345

Expected Output

Name:         my-app-12345

Namespace:    default

Containers:

  my-app:

    Requests:

      cpu:     100m

      memory:  200Mi

    Limits:

      cpu:     200m

      memory:  400Mi

Events:

  Type    Reason     Age   From               Message

  ----    ------     ----  ----               -------

  Normal  Scheduled  2m    default-scheduler  Successfully assigned default/my-app-12345 to node-1

This command lists Horizontal Pod Autoscalers, which automatically adjust the number of pods based on resource use. It helps you see if autoscaling is set up.

Terminal

kubectl get hpa

Expected Output

NAME       REFERENCE           TARGETS         MINPODS   MAXPODS   REPLICAS   AGE

my-app-hpa Deployment/my-app   50%/80%        1         5         2          10m

Key Concept

If you remember nothing else from resource monitoring, remember: always check both current usage and resource limits to keep your apps healthy and efficient.

Common Mistakes

Not setting resource requests and limits in pod specs.

Without these, Kubernetes can't manage resources well, leading to crashes or slow apps.

Always define CPU and memory requests and limits in your pod configuration.

Ignoring node resource usage and focusing only on pods.

Nodes can become overloaded even if individual pods look fine, causing cluster issues.

Monitor both pod and node resource usage regularly.

Not using autoscaling to adjust pod counts based on load.

Without autoscaling, your app may be underpowered during traffic spikes or waste resources when idle.

Set up Horizontal Pod Autoscalers to scale pods automatically based on CPU or memory use.

Summary

Use 'kubectl top pods' and 'kubectl top nodes' to see current resource usage.

Check pod resource requests and limits with 'kubectl describe pod'.

Use 'kubectl get hpa' to verify autoscaling is configured.

Set resource requests and limits to help Kubernetes manage your apps well.