



- Use
kubectlto change contexts between clusters. - SSH to control-plane or worker nodes when an individual task requires node-level access.
The student node is a client shell where you start. To access control-plane or worker nodes directly, use SSH as directed by individual tasks.

- Click the lab to start. Typical load time is ~30 seconds, but sometimes it may take 1–2 minutes.
- Once loaded, the question list appears (usually ~20 questions for this lab).
cluster3 and decode a Secret named beta-sec-cka14-arch located in the beta-ns-cka14-arch namespace. Follow these steps:
- Confirm which clusters are configured.
- Switch context to
cluster3. - Verify nodes in the selected cluster.
- Inspect the Secret, decode the base64 value, and save the decoded secret to a file on the student node such as
/opt/beta-sec-cka14-arch.
cluster3:
cluster1 context):
cluster3 you might see a single-node control plane (for example, cluster3-controlplane). Always set the correct context per question — many tasks will explicitly tell you which cluster to use.
Inspecting and decoding the Secret
First, list namespaces to confirm the target namespace exists:
- The exam interface allows scrolling between questions. Use the question list to jump back and forth.
- Each question belongs to one domain (architecture, scheduling, networking, storage, troubleshooting).
- Troubleshooting usually forms the largest share of questions (~30%).
- Complete the exam within the allotted time. If time expires, the exam is automatically submitted and validated.
- You can manually click End Exam to finish early and trigger validation.
Manage your time carefully. If the timer hits zero the exam will be auto-submitted and validated immediately.
kubectl context or using --context on commands. Others require SSH to control-plane or worker nodes for node-level changes.
Example — create a service account with limited permissions on cluster1:
- Read each task carefully to confirm which cluster and namespace to use.
- Use
kubectl --context <context>when working across clusters to avoid accidental changes. - Practice common troubleshooting patterns: logs, events, describing pods, and node-level checks.
- Familiarize yourself with file locations on the student node (e.g.,
/opt) when tasks require saving outputs.