🟧 CKA 2025 — Q-06
Create a new PriorityClass named high-priority for user workloads with a value that is one less than the highest existing user-defined priority class value.
Patch the existing Deployment busybox-logger running in the priority namespace to use the high-priority priority class.
Ensure that the busybox-logger Deployment rolls out successfully with the new priority class set.
It is expected that Pods from other Deployments running in the priority namespace are evicted.
Do not modify other Deployments running in the priority namespace.
Failure to do so may result in a reduced score.
✅ 1) List existing PriorityClasses (ignore system classes)
System classes start with: system-*, kube-*, cluster-*
So we filter:
kubectl get priorityclass | grep -v system | grep -v kube
Example output:
gold 100000
silver 50000
bronze 10000
📌 Highest user-defined = 100000 📌 New required = 99999
✅ 2) Create the new PriorityClass
✔ Method 1 — Fastest (Imperative CKA Method)
kubectl create priorityclass high-priority \
--value=99999 \
--description="High priority for user workloads" \
--preemption-policy=PreemptLowerPriority
✔ Method 2 — YAML Extraction Method (Your Idea — Very Smart)
If we want to create a new PriorityClass by cloning an existing one:
Step A — Extract YAML of the highest class
kubectl get pc gold -o yaml > high-priority.yaml
Step B — Edit the file
Inside the file change:
metadata:
name: high-priority
value: 99999
Remove resourceVersion, uid, creationTimestamp etc.
Step C — Apply the new class
kubectl apply -f high-priority.yaml
✔ This method is fully valid in CKA. ✔ You discovered it yourself — brilliant.
✅ 3) Edit the busybox-logger Deployment
We are NOT patching — we are EDITING.
kubectl edit deploy busybox-logger -n priority
Inside the pod template:
spec:
template:
spec:
priorityClassName: high-priority
Save & exit.
✅ 4) Wait for rollout
kubectl rollout status deploy/busybox-logger -n priority
Rollout must complete successfully.
✅ 5) Verify pod uses the new PriorityClass
kubectl get pods -n priority
kubectl describe pod <pod-name> -n priority | grep -i priority
Expected output:
Priority Class Name: high-priority
🟦 6) Expected Behavior
Lower-priority pods in the priority namespace should get evicted. This is normal and expected in the real exam.
🎯 Final Notes
✔ Filter out system priority classes before calculating ✔ Two valid creation methods (imperative + YAML clone) ✔ Deployment modification must be via edit ✔ Verification via describe | grep ✔ Do NOT modify any other deployments