Score above 70% to open Section 2.
Before you start
This is one 30-minute section of CKAD · Mission 01 · Sillage (16 weighted questions, 100 points). The timer starts when your terminal is ready. Closing the tab does not pause it. Score above 70% on this section to unlock the next.
You are working the decks of the corvette Sillage (namespaces hangar, cargo, workshop, command…). The Ricochet shuttle is docked under the belly. Other mocks share this cluster; change only what a question names.
Navigating questions
Use the Question dropdown above to open each statement and see its weight. Flag a question with ⚑ Flag for later review. Move between questions with ← / →.
In the terminal:
k — alias for kubectl
helm — available offline; repo station-charts is preconfigured
Answer and work files go under /opt/course/d/<n>/… as stated in each question.
Environment
- No outbound Internet — images, CoreDNS, and Helm charts are preloaded offline.
- CoreDNS is available — Service DNS names and ClusterIPs both work.
- No Docker daemon — you will not build images.
- Helm is offline. Search charts with
helm search repo station-charts/.
- Preloaded images:
nginx:1.14.2, busybox:1.36.
Scoring
The live cluster is scored automatically every 5 minutes. Open View score / rubric to see the breakdown, or click Grade now for an immediate check. You can grade again after fixing tasks. Your latest score is saved to your account.
Tips
- Answer questions in any order — weights are shown in the dropdown.
- 30 minutes, no pause — manage your time.
- Partial credit is awarded per subtask.
Question 1 (6 points) — InitContainer, Volumes
Create a Pod ricochet in namespace flight:
- a container shuttle, image nginx:1.14.2
- an init container preflight, image busybox:1.36, that creates the
file /shared/ready before shuttle starts
- both containers mount the same emptyDir volume at /shared
Question 2 (7 points) — Job
Create a Job timonerie-run in namespace hangar:
- one container runner, image busybox:1.36, running the command
echo timonerie-ok
- the Job needs 3 successful completions, with at most 2 Pods running
in parallel
- the Job retries at most 2 times before it is marked failed
Question 3 (6 points) — CronJob
Create a CronJob boulon-tune in namespace workshop:
- schedule */5 * * * * (every 5 minutes)
- a new run must not start while the previous one is still running;
skip it instead
- only the last successful Job is kept in history
- one container droid, image busybox:1.36, running the command
echo tune
Question 4 (6 points) — Volumes
Create a Pod spray-paint in namespace workshop with one container
sorrel, image busybox:1.36, command sleep 1d.
Add an emptyDir volume backed by memory (tmpfs) instead of the node's
disk, and mount it in sorrel at /paint.