Kubernetes, EKS, GitOps & releases
Helm & Kustomize
Helm and Kustomize are the two standard ways to package and customise Kubernetes manifests, and almost every Senior DevOps interview asks about at least one of them. Many teams use Kustomize with Flux day to day, but interviewers also expect Helm fluency — chart structure, templating, values precedence, releases and rollbacks, hooks, OCI registries, chart testing — and a clear view of when to use which (and how to combine them).
After this topic you can
- Package an app as a Helm chart with values per environment
- Patch manifests with Kustomize bases and overlays
- Choose between Helm and Kustomize for a team
Lessons
- 1The problem both tools solve
-
2Helm fundamentals
- Concepts
- Chart layout
- Everyday commands
- Values precedence
-
3Templating in depth
- The basics
- Named templates (_helpers.tpl)
- Library charts and an internal "service chart"
-
4Releases, hooks, tests and operations
- How upgrades work
- Hooks
- CRDs
- OCI registries
- Testing and quality in CI
- Secrets with Helm
-
5Kustomize in depth
- Base and overlays
- Transformers and generators
- Kustomize with Flux (a typical setup)
-
6Choosing and combining
- Decision guide
- Pitfalls to mention in interviews
-
7Interview questions and model answers
- "Helm vs Kustomize — when do you use which?"
- "Explain Helm values precedence."
- "An upgrade failed and Helm says another operation is in progress."
- "How do you make pods restart when a ConfigMap changes?"
- "How do you manage secrets with Helm/Kustomize in GitOps?"
- "How do you test charts in CI?"
- "What happens to CRDs in a chart's crds/ folder on upgrade?"