Kubernetes weekend roadmap + interview tips
A practical weekend roadmap covers Pods, Deployments, Ingress, Secrets, probes, and Helm charts for beginners—paired with interview tips that emphasize cluster-scale and limits questions. It’s a compact way to convert hands-on practice into interview-ready talking points on real-world K8s management. (x.com) (x.com)
Kubernetes defines a Pod as the smallest deployable unit that groups one or more containers with shared storage and network resources. (kubernetes.io) A Deployment provides declarative updates for ReplicaSets and is the standard controller for rolling updates, rollbacks, and scaling stateless workloads. (kubernetes.io) Kubernetes supports three probe types—startup, liveness, and readiness—and the kubelet uses liveness probes to restart deadlocked containers while readiness probes remove unhealthy Pods from service endpoints. (kubernetes.io) Ingress is the API object that maps HTTP(S) traffic to Services and is implemented by ingress controllers such as ingress-nginx, which provides Helm chart installation and manifests for cluster-specific deployment. (kubernetes.io, kubernetes.github.io/ingress-nginx) Helm packages applications into charts (Chart.yaml, values.yaml, templates/) to enable versioned installs, in-place upgrades, and rollbacks, and the Helm project documentation and Chart Template Guide walk through chart authoring and the Helm v4 toolchain. (helm.sh, helm.sh) Production guidance for secrets with Helm emphasizes encrypted-in‑Git tools like SOPS with the helm‑secrets plugin or syncing secrets from external stores via External Secrets Operator or Vault rather than storing plaintext Kubernetes Secrets. (blog.gitguardian.com, github.com/jkroepke/helm-secrets, hariknine.github.io) Interview-prep guides now push cluster-scale topics—expect questions on resource requests/limits, Horizontal Pod Autoscaler behavior, API traffic under scale, and operational tradeoffs—while official Kubernetes limits document cites capacity ceilings used as anchors in those discussions (Kubernetes v1.35: up to 5,000 nodes, ~150,000 total pods, and recommended max ~110 pods per node). (datacamp.com, educative.io, kubernetes.io)
