Kubernetes Overview

First PublishedFeb 16, 2026Last UpdatedMay 18, 2026ByAtif Alam

Kubernetes (often shortened to K8s) is an open-source container orchestration platform.

It automates deploying, scaling, and managing containerized applications across clusters of machines.

Why Kubernetes?

• Containers alone aren’t enough. A single docker run works for one machine, but production needs scheduling across many nodes, self-healing, load balancing, rolling updates, and secret management.
• Kubernetes handles that orchestration. You describe the desired state (e.g. “run 3 replicas of this container”) and Kubernetes continuously works to make reality match.

Key Terminology

• Cluster — A set of machines (nodes) managed by Kubernetes.
• Node — A single machine in the cluster (physical or virtual). Runs containers.
• Pod — The smallest deployable unit; one or more containers that share network and storage.
• Service — A stable network endpoint that routes traffic to a set of pods.
• Deployment — Declares the desired state for pods (image, replicas, update strategy). Kubernetes creates and manages the pods.
• Namespace — A virtual partition inside a cluster for isolating resources.
• kubectl — The CLI tool for interacting with a Kubernetes cluster.

Topics in This Section

Order matches the sidebar: foundations → networking → storage and workloads → delivery → production → security → review pages → examples and EKS.

• Architecture — Control plane, worker nodes, and how the pieces fit together.
• Scheduling and placement — Predicate vs priority scheduling, PriorityClass, and Pending triage vocabulary.
• etcd and control plane health — etcd role, health signals, snapshots, and control-plane certificate rotation.
• Core Objects — Pods, Deployments, Namespaces, Labels, and YAML anatomy.
• Manifests — YAML structure, the four top-level fields, spec vs status, and applying manifests.
• Networking — Services, Ingress, Gateway API vs Ingress, CNI comparison, DNS, network policies, and CoreDNS failure modes.
• Services and endpoints — ClusterIP datapath, EndpointSlices, kube-proxy modes, and debugging.
• Network policies — Default deny, allowlists, metadata egress, and CNI compatibility.
• Sidecar Pattern — When to use sidecars, sidecar vs library trade-offs, and production rollout patterns.
• Ingress Controllers — NGINX, Traefik, and AWS Load Balancer Controller with TLS/mTLS and cert-manager patterns.
• Istio — Service mesh architecture, VirtualService, mTLS, and troubleshooting with istioctl.
• Storage — Volumes, PVCs, StorageClass binding modes, topology, and Secrets.
• Workload Types — Deployments vs StatefulSets vs DaemonSets vs Jobs and CronJob patterns.
• Stateful backup and restore — Snapshots, Velero, restore drills, and RPO/RTO.
• Helm — Package management for Kubernetes.
• Helm Templating — Go template syntax, value injection methods, and where values come from in production.
• Helm vs operators vs GitOps — When to use Helm charts, operators, and Argo CD/Flux, and how they work together.
• Production Platform Checklist — Layered platform checks for ownership, blast radius, delivery guardrails, and drift.
• Production Patterns — Health checks, resource limits, noisy-neighbor mitigation, PDBs, draining, and rolling updates.
• Autoscaling on EKS — HPA, VPA, KEDA, Cluster Autoscaler, and Karpenter; custom metrics via Prometheus or CloudWatch.
• Production Scenarios — Scenario-based practice for production reasoning and mitigation planning.
• Cluster upgrades — Control plane vs nodes, add-on smoke tests, CRD compatibility, and certificate inventory.
• Operators — CRDs, custom controllers, the reconcile loop, and building your own operator.
• Kubectl Reference — Common commands grouped by task.
• Kubeconfig and authentication — Client TLS, OIDC, and EKS get-token patterns.
• RBAC — Roles, bindings, blast radius, and kubectl auth can-i.
• Pod Security Standards — What PSA blocks vs allows, choosing Baseline or Restricted, rollout, exemptions, and Restricted-friendly manifests.
• Admission controllers — Mutating/validating webhooks, failurePolicy, and policy engines.
• Multi-tenancy and policy — Namespaces, quotas, shared ingress fairness, and audit logging.
• Multi-cluster management — Cluster API, Karmada, and GitOps fleet patterns across clusters.
• Troubleshooting and Debugging — A practical production triage flow across workloads, cluster signals, and networking.
• Kubernetes architecture review questions — High-signal prompts for architecture reviews, hiring calibration, and production readiness checks.
• Kubernetes architecture review answers — Concise reference answers for the discussion prompts by domain, with links to deeper guides.
• Examples — File-oriented layouts (base workloads, Istio, kubectl vs Argo CD).
• EKS (AWS) — Amazon EKS overview and a production-oriented cluster with Terraform (VPC, private API, node groups, add-ons). Uses the AWS networking and VPC connectivity guides for prerequisites.
• Migrating workloads from EC2 to EKS — Phased cutover, DNS and traffic shift, data, and rollback framing.
• Prometheus Adapter for HPA on EKS — Custom metrics from PromQL for HPA.
• Container Insights for HPA on EKS — CloudWatch external metrics for HPA.
• EKS troubleshooting cheat sheet — Symptom-driven EKS debugging runbook for networking, autoscaling, and on-call triage.