Skip to content

Level 3: Production Kubernetes

Core K8s objects, probes, rollouts, services, DNS, scaling.

Concepts

pod, deployment, service, endpoints, namespace, labels, configmap, secret, readiness_probe, liveness_probe, horizontal_pod_autoscaler, resource_limits, resource_requests, rollout, replicaset, dns, networkpolicy, ingress, rbac, statefulset, daemonset, persistent_volume, pod_disruption_budget

Failure Patterns You Should Be Able to Resolve

  • FP-001: crashloopbackoff
  • FP-002: readiness_probe_failure
  • FP-003: imagepullbackoff
  • FP-004: oomkilled
  • FP-005: hpa_not_scaling
  • FP-006: dns_resolution_failure
  • FP-007: networkpolicy_block
  • FP-008: service_selector_mismatch
  • FP-012: rollout_stuck
  • FP-013: ingress_404
  • FP-014: rbac_forbidden

Commands You Should Be Fluent With

  • kubectl get pods -n <ns> / -o wide / --show-labels
  • kubectl logs -n <ns> deploy/<name> / --previous
  • kubectl describe pod -n <ns> <pod>
  • kubectl get endpoints -n <ns>
  • kubectl rollout status / undo / history
  • kubectl get hpa / describe hpa
  • kubectl top pods / top nodes
  • kubectl exec -n <ns> <pod> -- <cmd>
  • kubectl run <debug> --rm -i --restart=Never --image=busybox -- <cmd>
  • kubectl auth can-i
  • kubectl get events --sort-by=.lastTimestamp

Assets to Complete

Read (theory)

Practice (hands-on)

Runbooks (study)

Review (flashcards)