Skip to content

Latest commit

 

History

History
200 lines (162 loc) · 4.3 KB

Lab-Pod-Probes-Debugging.md

File metadata and controls

200 lines (162 loc) · 4.3 KB

Kubernetes Probes & Debugging Lab

1. Create a Pod with an HTTP Startup Probe

Step 1: Generate a Pod YAML Manifest

We will create a Pod named web-server using the nginx image and expose port 80. Use the kubectl run command in dry-run mode to generate the YAML file:

kubectl run web-server --image=nginx --port=80 --restart=Never \
-o yaml --dry-run=client > probed-pod.yaml

Step 2: Add a Startup Probe

Edit the probed-pod.yaml file to include a startup probe that performs an HTTP GET request on the root (/) path of the container:

apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web-server
  name: web-server
spec:
  containers:
    - image: nginx
      name: web-server
      ports:
        - containerPort: 80
          name: nginx-port
      startupProbe:
        httpGet:
          path: /
          port: nginx-port
  dnsPolicy: ClusterFirst
  restartPolicy: Never

2. Add a Readiness Probe

Modify the probed-pod.yaml to include a readiness probe. This probe ensures that the container is ready to accept traffic. It should:

  • Perform an HTTP GET request on /.
  • Wait 5 seconds before the first check.
readinessProbe:
  httpGet:
    path: /
    port: nginx-port
  initialDelaySeconds: 5

Updated probed-pod.yaml:

apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web-server
  name: web-server
spec:
  containers:
    - image: nginx
      name: web-server
      ports:
        - containerPort: 80
          name: nginx-port
      startupProbe:
        httpGet:
          path: /
          port: nginx-port
      readinessProbe:
        httpGet:
          path: /
          port: nginx-port
        initialDelaySeconds: 5
  dnsPolicy: ClusterFirst
  restartPolicy: Never

3. Add a Liveness Probe

Modify probed-pod.yaml to include a liveness probe. This probe ensures that the container is still running properly. It should:

  • Perform an HTTP GET request on /.
  • Wait 10 seconds before the first check.
  • Check every 30 seconds.
livenessProbe:
  httpGet:
    path: /
    port: nginx-port
  initialDelaySeconds: 10
  periodSeconds: 30

Final probed-pod.yaml:

apiVersion: v1
kind: Pod
metadata:
  labels:
    run: web-server
  name: web-server
spec:
  containers:
    - image: nginx
      name: web-server
      ports:
        - containerPort: 80
          name: nginx-port
      startupProbe:
        httpGet:
          path: /
          port: nginx-port
      readinessProbe:
        httpGet:
          path: /
          port: nginx-port
        initialDelaySeconds: 5
      livenessProbe:
        httpGet:
          path: /
          port: nginx-port
        initialDelaySeconds: 10
        periodSeconds: 30
  dnsPolicy: ClusterFirst
  restartPolicy: Never

4. Deploy and Monitor the Pod

Step 1: Create the Pod

kubectl create -f probed-pod.yaml

Step 2: Verify Pod Status

kubectl get pod web-server

The output should show the container transitioning from ContainerCreating to Running.

Step 3: Inspect Probe Configurations

kubectl describe pod web-server

Step 4: Retrieve Pod Metrics

Use the Kubernetes Metrics Server to check CPU and memory usage:

kubectl top pod web-server

5. Debugging a Failed Pod

Step 1: Create a Pod with a Custom Command

We will create a Pod named custom-cmd using the busybox image. The container will attempt to run top-analyzer --all, which does not exist in the image:

kubectl run custom-cmd --image=busybox --restart=Never \
-- /bin/sh -c "top-analyzer --all"

Step 2: Check Pod Status

kubectl get pod custom-cmd

The output will show the Pod in Error state.

Step 3: Troubleshoot the Pod

Check the logs for failure messages:

kubectl logs custom-cmd

The output will indicate that top-analyzer is not available in the busybox image.

Summary

  • We created a Pod with probes to ensure proper startup, readiness, and liveness.
  • We monitored the Pod’s lifecycle and resource usage.
  • We troubleshot a failed Pod due to a missing command.

This lab helps in understanding Kubernetes probes, monitoring, and debugging in real-world scenarios. 🚀