System Design: Practical Kubernetes (K8s) Use and Operations

Context: In a technical screen, you are asked to describe how you have used Kubernetes in production to build and operate services. Provide concrete practices, tools, and trade-offs you made.

Cover the following areas:

1. Deployments and rollout strategies
   • How you package manifests (kubectl, Helm, Kustomize) and promote across environments
   • Readiness/liveness/startup probes, resource sizing, PDBs, surge/unavailable settings
   • GitOps or CI/CD approach; blue/green or canary
2. Services and Ingress
   • Service types (ClusterIP/LoadBalancer), internal vs. external routing
   • Ingress controller choice, TLS/certificate management, host/path routing
   • Service mesh usage (if any) and when it’s helpful
3. Autoscaling
   • HPA/VPA/Cluster Autoscaler configuration and metrics used
   • Stabilization, min/max bounds, and preventing thrash
4. Configuration and secret management
   • ConfigMaps vs. Secrets, mounting patterns, reload strategies
   • Encryption, rotation, and access controls (RBAC)
5. Observability
   • Metrics, logs, traces; dashboards and alerting
   • Golden signals and SLOs; synthetic checks
6. Rollback and upgrade strategies
   • Application rollbacks and database migrations
   • Node/cluster upgrades, draining, PDBs, and disruption management
7. Specific challenges and how you solved them
   • Share 2–3 concrete issues (e.g., downtime during rollout, HPA flapping, NAT exhaustion) and your resolution