Overview
Application engineers practice shipping manifests with guardrails: resource limits, rollout strategies, and Jobs that behave when clusters are noisy. Labs emphasize readability and diff-friendly YAML so reviews in your organization stay painless.
What you work through
- Probe tuning labs with failing health checks on purpose
- Init container sequencing with realistic sidecar patterns
- ConfigMap and Secret rotation drills without downtime
- CronJob failure recovery with backoff interpretation
- NetworkPolicy authoring from plain-language requirements
- HPA wiring with custom metrics stubs
- Git-friendly packaging patterns for Helm-free teams
Outcomes
- Author a multi-container Pod manifest with justified probes
- Explain how you would roll back a failed Deployment safely
- Complete a CKAD-style sprint with readable YAML comments
Facilitator
Sora Kim
DevOps Curriculum Designer specializing in developer-centric Kubernetes workflows.
Participant notes
The probe failure lab mirrored what we see in staging. I appreciated the blunt note that noisy metrics can mask real failures.
ConfigMap rotation section felt compressed, yet the office hour walkthrough clarified the edge cases.
Course questions
No. We focus on core workload APIs. Optional appendix covers Helm only if your team already adopted it.
Labs receive narrative feedback, not numeric grades. We highlight risk areas instead of ranking learners.
IDE licenses, private cluster sandboxes beyond the cohort window, and CKAD exam fees.