Member-only story
5 Easy Tips for Troubleshooting Your Kubernetes Pods
Troubleshoot your Kubernetes application with ease and understand its mechanics
In many cases, you may find that an application in Kubernetes has not been deployed correctly or is not working as it should. This post offers troubleshooting tips to resolve such situations swiftly.
After reading this piece, you’ll also gain insights into the internals of Kubernetes, and as a bonus, I’ll share helpful tips on navigating K8S on your own.
So, let’s begin!.
First, there are two reasons why your Pods can fail.
- Errors in the configuration of your Kubernetes resources like deployment and services.
- Problems in your code.
In the former case, containers do not start. In the latter instance, the application code fails after the container starts up. We’ll address each of these situations systematically.
The kubectl command-line utility will be used to interact with K8S throughout this exercise.
Tip 1: Observe Pods
Verify that the Pods are in status Running or Ready.
kubectl get pods
One Pod is in the status Pending for nine hours, that cannot be good! The container did not start, and we’ll investigate this with the describe command in tip number two.
Meanwhile, here are other error codes that occur when a container fails to start.
ImagePullBackoff— Docker image registry not accessible, image name/version specified in deployment incorrect. Make sure the image name is correct, and the registry is accessible and authenticated (docker login…).RunContainerError— One possible cause:ConfigMap/Secretsmissing.ContainerCreating— Something not available immediately, persistent volume?
Before we investigate other errors, let’s experiment with starting a Pod using an incorrect image name.
