Skip to content

README.md

Latest commit

 

History

History
201 lines (139 loc) · 3.59 KB

README.md

File metadata and controls

201 lines (139 loc) · 3.59 KB

Streamline your troubleshooting

In this lab, you'll learn how to troubleshoot common issues in Kubernetes applications in a streamlined manner by:​

  1. Identifying the problem
  2. Gathering relevant information
  3. Analyzing logs and events
  4. Diagnosing the root cause
  5. Applying fixes and verifying the solution​

There's also a Talk at Cloud Native Zurich about this topic. Feel free to check it out in your own time.

Scenario 1: Pod CrashLoopBackOff

📝 Step 1: Deploy a Faulty Application

Apply the following manifest to deploy an application that will crash:​

apiVersion: v1
kind: Pod
metadata:
  name: crashpod
spec:
  containers:
    - name: crashcontainer
      image: busybox
      command: ["sh", "-c", "exit 1"]

Save this as crashpod.yaml and apply it:​

kubectl apply -f crashpod.yaml

🔍 Step 2: Identify the Issue

Check the pod status:​

kubectl get pods

📄 Step 3: Gather Logs

Retrieve the logs to understand why the container is crashing:​

kubectl logs crashpod

Since the container exits immediately, you might need to check previous logs:​

kubectl logs crashpod --previous

🛠️ Step 4: Diagnose and Fix

The container is exiting with status code 1. Modify the command to keep the container running:​

command: ["sh", "-c", "while true; do echo Hello; sleep 10; done"]

Update the pod:​

kubectl delete pod crashpod
kubectl apply -f crashpod.yaml

Verify the pod is running:​

kubectl get pods

🔍 Scenario 2: ImagePullBackOff

📝 Step 1: Deploy with Incorrect Image

Apply a manifest with a non-existent image:​

apiVersion: v1
kind: Pod
metadata:
  name: badimagepod
spec:
  containers:
    - name: badimage
      image: nonexistentrepo/nonexistimage:latest

Save this as badimagepod.yaml and apply it:​

kubectl apply -f badimagepod.yaml

🔍 Step 2: Identify the Issue

Check the pod status:​

kubectl get pods

The pod should be in an ImagePullBackOff state.​

📄 Step 3: Describe the Pod

Get detailed information about the pod:​

kubectl describe pod badimagepod

Look for events indicating image pull errors.​

🛠️ Step 4: Fix the Image Reference

Update the image to a valid one, such as nginx:latest. Modify the manifest accordingly and reapply:​

kubectl delete pod badimagepod
kubectl apply -f badimagepod.yaml

Verify the pod is running:​

kubectl get pods

🔍 Scenario 3: Pending Pod Due to Resource Constraints

📝 Step 1: Deploy a Pod with High Resource Requests

Apply the following manifest:​

apiVersion: v1
kind: Pod
metadata:
  name: highresourcepod
spec:
  containers:
    - name: highresource
      image: nginx
      resources:
        requests:
          memory: "10Gi"
          cpu: "4"

Save this as highresourcepod.yaml and apply it:​

kubectl apply -f highresourcepod.yaml

🔍 Step 2: Identify the Issue

Check the pod status:​

kubectl get pods

The pod should be in a Pending state.​

📄 Step 3: Describe the Pod

Get detailed information:​

kubectl describe pod highresourcepod

Look for messages indicating insufficient resources.​

🛠️ Step 4: Adjust Resource Requests

Reduce the resource requests to fit within the cluster's capacity:​

resources:
  requests:
    memory: "256Mi"
    cpu: "0.5"

Update the manifest and reapply:​

kubectl delete pod highresourcepod
kubectl apply