Kubernetes Interview Questions and Answers & Incident Runbooks (2026 Edition)
⚡ 207 Live Scenarios
🎯 STAR Method Answers📋 60s Elevator Pitches
Kubernetes has become the de facto operating system for cloud-native infrastructure, making deep container orchestration expertise essential for senior platform engineers and Site Reliability Engineers. In production environments, mastering Kubernetes goes far beyond writing standard Deployment manifests or running basic imperative kubectl commands. Candidates facing rigorous technical interview loops are expected to diagnose complex failure modes such as CrashLoopBackOff cascades, ImagePullBackOff authentication failures, OOMKilled evictions triggered by Linux cgroup memory limits, and silent etcd quorum loss during network partitions. Furthermore, interviewers probe candidate intuition regarding control plane reconciliation loops, scheduler scoring and binding algorithms, custom admission controller webhooks, and advanced CNI networking nuances like Cilium eBPF packet routing versus legacy iptables kube-proxy overhead. Our comprehensive collection of Kubernetes scenario questions provides step-by-step diagnostic runbooks, production-grade CLI commands, and proven STAR-framework architectural narratives. Learn how to debug core DNS latency spikes, rightsize vertical and horizontal autoscalers (HPA and VPA), architect zero-downtime rolling updates with pod disruption budgets, and implement hardened Kyverno or OPA Gatekeeper governance across multi-tenant enterprise clusters.
Get 1 Kubernetes interview question in your inbox every week
Preparing for a cloud or platform interview? Subscribe to get our weekly deep-dive scenario plus instant access to the Top 50 Kubernetes Interview Questions & Incident Runbooks PDF.
Prefer video walkthroughs? Watch Naveed break down real-world Kubernetes troubleshooting scenarios—from CrashLoopBackOff and OOMKilled pods to Ingress 502 diagnostics and CNI failures—with candidate storytelling strategies and architectural STAR answers.
Production runbook strategy for upgrading Amazon EKS clusters across minor versions (v1.34 → v1.35 → v1.36 → v1.37) with zero application downtime using sequent...
Fast-response production incident runbook for rolling back broken Kubernetes deployments across native kubectl, GitOps (ArgoCD/Flux), and handling database migr...
Exhaustive diagnostic methodology for intermittent Kubernetes failures: node kernel pressure, CPU throttling, CoreDNS latency (ndots:5), CNI IP exhaustion, and ...
Systematic runbook to troubleshoot ImagePullBackOff in Kubernetes: isolate tag typos, verify registry credentials, test VPC network egress, and inspect containe...
Definitive architectural guide for configuring CPU and memory requests/limits: how kube-scheduler uses requests, how the Linux kernel enforces limits, CFS CPU t...
Deep dive into Kubernetes HPA: metrics collection pipeline via Metrics Server and Custom Metrics API, the exact mathematical autoscaling formula, stabilization ...
End-to-end investigative procedure for identifying and resolving runaway CPU and memory saturation in Kubernetes clusters: isolating nodes vs pods, diagnosing m...
Check pod labels vs service selector — kubectl describe service shows the selector. kubectl get pod --show-labels shows pod labels. They must match exactly....
Debug a stuck Kubernetes init container by inspecting pod events with describe, checking init container logs, and verifying blocking service dependencies....
A DaemonSet ensures one pod runs on every node (or a subset of nodes). When a new node joins the cluster, the DaemonSet automatically places a pod on it....
PV (PersistentVolume) — the actual storage resource. Could be an AWS EBS volume, NFS share, local disk. Created by a cluster admin or dynamically provisioned....
Yes — this is a supported temporary state during upgrades. Kubernetes supports N-2 version skew between control plane and nodes. A 1.27 control plane can manage...
A PodDisruptionBudget (PDB) limits how many pods of a deployment can be voluntarily disrupted at the same time. "Voluntary disruption" includes node d...
A LimitRange sets default and maximum resource requests/limits for pods in a namespace. If a pod doesn't specify resources, LimitRange fills in defaults....
Liveness probe — "is this container still alive?" If it fails, Kubernetes restarts the container. Use for detecting deadlocks or infinite loops where ...
emptyDir — temporary directory created when pod starts, deleted when pod is removed. Shared between containers in the pod. Good for scratch space or sidecar dat...
Pod is unschedulable for a reason other than resources — e.g., node affinity requires a specific label that no node type has. CA won't add nodes it can...
Helm is a package manager for Kubernetes. A Helm chart bundles all the Kubernetes YAML for an application (Deployment, Service, ConfigMap, Ingress, etc.) with t...
When a pod is deleted, Kubernetes sends SIGTERM to the container and waits terminationGracePeriodSeconds (default: 30s) for the app to shut down gracefully. Aft...
This spreads pods across availability zones rather than just across nodes. If your cluster spans 3 AZs (us-east-1a, 1b, 1c), this anti-affinity ensures no two p...
Admission webhooks intercept API requests before they're stored in etcd. Mutating webhooks can modify the request (add/change fields). Validating webhooks ...
Service mesh vs NetworkPolicy is not an either-or choice: NetworkPolicy is a built-in L3/L4 firewall enforced by the CNI that decides which pods may connect, wh...
A pod restart on config change is not automatic: Kubernetes only rolls pods when the Pod template changes, so you must make the config change show up in the tem...
kube-proxy implements Services using iptables by watching Services and EndpointSlices and programming NAT rules that rewrite a Service's virtual ClusterIP ...
Topology spread constraints tell the scheduler to distribute matching pods evenly across failure domains such as zones or nodes, within a tolerated difference c...
When a pod needs GPU resources, you configure it by running a device plugin on GPU nodes and requesting the extended resource, for example nvidia.com/gpu: 1, in...
Leader election in the Kubernetes control plane lets several replicas of kube-controller-manager and kube-scheduler run for high availability while only one act...
The Kubernetes garbage collector is a controller in kube-controller-manager that deletes dependent objects once their owners are gone, using metadata.ownerRefer...
The Kubernetes watch mechanism is a long-lived HTTP request that streams change events (ADDED, MODIFIED, DELETED) for resources starting from a given resourceVe...
kubectl apply -f vs apply -k: -f applies the manifest files exactly as written, while -k first builds a directory containing a kustomization.yaml (adding overla...
Snapshot PVC data (Velero). 2) Deploy workload in new cluster from same Git source. 3) Restore data snapshot to new cluster PVCs. 4) Test new cluster. 5) Switch...
To expose a gRPC service in Kubernetes, put it behind a Service for in-cluster traffic and an HTTP/2-aware entry point (an Ingress or Gateway API route with TLS...
Same as any pod: kubectl describe job , check the created pod's events. Common issues: image pull error, no nodes with enough resources, node selector mism...
The control loop pattern: watch current state → compare with desired state → take action to reconcile. Controllers (Deployment controller, ReplicaSet controller...
kube-proxy runs on every node (as a DaemonSet) and maintains network rules (iptables or ipvs) that implement Services. It watches the API server for Service/End...
In iptables mode: kube-proxy creates a chain of iptables rules for each Service. With thousands of Services, rule traversal becomes linear (O(n)). Performance d...
Node is under resource pressure. kubelet evicts lower-priority pods to free resources. Check: kubectl describe pod — reason will say Evicted with a reason (memo...
Architecture and networking patterns for Kubernetes multi-cluster service discovery: evaluating the Kubernetes MCS API (ServiceExport/ServiceImport), Cilium Clu...
To configure resource requests and limits for init containers, set resources on each init container just like an app container, but remember the scheduler compu...
A projected volume in Kubernetes maps several existing volume sources into one directory, so a container sees a ConfigMap, a Secret, downward API fields and a S...
Deep dive into the Kubernetes Downward API: fieldRef vs resourceFieldRef, injecting pod/container metadata into environment variables and downwardAPI volume mou...
Set PodDisruptionBudgets on all critical workloads. kubectl drain --ignore-daemonsets --delete-emptydir-data. The drain respects PDBs — won't proceed if it...
Production guide to Kubernetes VPA: Recommender, Updater, and Admission Controller components, deployment modes (Off, Initial, Recreate, Auto), and preventing f...
Three levels: Privileged (no restrictions), Baseline (blocks most dangerous capabilities — no privileged, no hostPath), Restricted (most secure — non-root, no c...
To get events for a specific namespace sorted by time, run kubectl get events -n prod --sort-by=.lastTimestamp, or use the newer kubectl events -n prod, which s...
When you delete a Kubernetes namespace, the API server marks it Terminating and the namespace controller deletes every namespaced object inside it, then removes...
Control plane nodes are tainted with node-role.kubernetes.io/control-plane:NoSchedule. Add a toleration to your pod: tolerations: [{key: "node-role.kuberne...
By default, K8s has NO network isolation. All pods can talk to all other pods in the cluster, across namespaces. This is intentional (for ease of use) but dange...
A helper container that runs alongside the main app container in the same pod, sharing its network and volumes. Examples: Istio proxy (Envoy), Fluentd for log s...
Federation v1 (deprecated) tried to manage multiple clusters from a single control plane. It was complex and unreliable. Federation v2 (KubeFed) also proved dif...
To create a self-signed TLS certificate for an Ingress, generate a key and certificate with openssl req -x509 including a SubjectAltName, store them as a kubern...
An admission controller is a piece of API server code that intercepts a request after authentication and authorization but before the object is persisted to etc...
kubectl apply --prune makes kubectl apply delete live objects that were previously applied by kubectl but are no longer present in the manifests you pass, so th...
Force all outbound traffic through a central point (useful for IP whitelisting at third-party APIs). With Istio: configure an EgressGateway service and VirtualS...
kubectl --dry-run=server vs client comes down to where the request is processed: with client kubectl builds and prints the object locally and never contacts the...
With Istio: EnvoyFilter or RateLimitPolicy using a Redis-backed rate limit service. With nginx-ingress: nginx.ingress.kubernetes.io/limit-rps annotation....
Architectural runbook for scaling Istio and Envoy service discovery across 1,000+ microservices without control-plane xDS push storms, memory bloat, or OOM cras...
How to design multi-tier Kubernetes liveness, readiness, and startup probes that detect internal thread deadlocks, downstream connection pool starvation, and da...
How to professionally articulate your current Kubernetes production version (v1.29/v1.30+), enforce supported version skew across control plane and worker nodes...
Real-world incident response narrative: diagnosing a sudden 5xx error spike caused by an invalid downstream endpoint in a ConfigMap feature flag, rolling back i...
Root-cause analysis and resolution for intermittent HTTP 502 Bad Gateway errors during traffic peaks: diagnosing upstream timeouts, CPU throttling, readiness pr...
Systematic triage when a Helm release fails due to resource constraints: diagnosing scheduling bottlenecks (node CPU/RAM starvation) vs namespace ResourceQuota ...
Modern strategy for publishing and sharing internal Helm charts across engineering teams via OCI registries (ECR, Harbor, GHCR) with semantic versioning, values...
Comprehensive automated testing pipeline for enterprise Helm charts: template linting, schema validation, Go template unit tests with helm-unittest, and integra...
Deep dive into what actually happens when a Kubernetes Pod 'restarts': differentiating in-place container restarts by Kubelet from Pod deletion/resche...
Deep architectural explanation of how kube-proxy programs node-level networking to route ClusterIP service traffic to backend Pods using iptables, IPVS, and mod...
Production runbook for safely evacuating workloads from unhealthy or degrading Kubernetes worker nodes without violating PodDisruptionBudgets or causing service...
Detailed breakdown of the Kubernetes health probe evaluation lifecycle, explaining why startupProbe runs first and how readiness and liveness interact to preven...
Comparing Kubernetes Operators vs Helm: Day-1 deployment packaging with templates vs Day-2 automated reconciliation loops, stateful failover, backups, and when ...
Deep dive into the Container Runtime Interface (CRI) in Kubernetes: gRPC client-server model, kubelet integration, containerd vs CRI-O, OCI runtime spec (runc),...
Architectural guide to Bin Packing in Kubernetes: understanding kube-scheduler node scoring plugins (MostAllocated vs LeastAllocated), descheduler defragmentati...
Analyzing the most difficult operational challenges of running Kubernetes at scale: cluster upgrade deprecations, networking complexity (MTU/DNS/CNI), storage d...
End-to-end troubleshooting methodology to diagnose and resolve intermittent liveness and readiness probe failures in Azure Kubernetes Service (AKS)....
Architecture blueprint for designing metric-driven automated rollback pipelines in Kubernetes using Argo Rollouts, Prometheus analysis templates, and deployment...
Comprehensive disaster recovery framework for stateful containerized applications running on Kubernetes, achieving strict RTO and RPO targets across cloud regio...
Deep dive into the operational differences between Kubernetes readiness, liveness, and startup probes, with a detailed breakdown of pod restart mechanics....
Deep architectural comparison between Kubernetes CPU Throttling and OOMKilled evictions, explaining how the Linux kernel enforces CPU CFS bandwidth versus memor...
The definitive master preparation guide for Kubernetes interview questions and answers: control plane internals (etcd, API server, scheduler), pod lifecycle sta...
Production incident triage for isolating and resolving worker nodes stuck in NotReady status across 5 failure domains: kubelet systemd process crashes, cgroup d...
Deep diagnostic analysis of why Kubernetes pods get stuck in Waiting state with Reason: CreateContainerError, dissecting volume attachment deadlocks (EBS Multi-...
Comprehensive guide to understanding and resolving CreateContainerConfigError, explaining how Kubelet validates environment variables, secretKeyRef, configMapKe...
Advanced production analysis of RunContainerError vs CrashLoopBackOff, focusing on missing entrypoint binaries, executable permission bits (chmod +x), scratch/d...
Master Kubernetes service discovery internals: examine how CoreDNS resolves cluster.local queries, EndpointSlice controllers track pod health, and kube-proxy co...
Best practices for Kubernetes ConfigMaps and Secrets across Dev, Staging, and Production environments: base64 pitfalls, External Secrets Operator (ESO), HashiCo...
Production guide to implement RBAC in Kubernetes: configure Roles, ClusterRoles, RoleBindings, ServiceAccounts, and enforce the principle of least privilege....
Key incident runbooks, interview talking points, and architecture tradeoffs.
Walk me through how you upgraded an Amazon EKS cluster from Kubernetes v1.34 to v1.36 and even after v1.37 without downtime.
In one of my projects, we had a customer-facing application running as microservices on Amazon EKS, and we had a requirement to upgrade Kubernetes from v1.34 to v1.36. Since it was Production, we couldn't afford application downtime, so we followed a proper upgrade runbook.
Key Architectural Takeaway: Zero downtime wasn't achieved just because we carefully upgraded EKS. It was possible because the application was already designed for high availability with multiple replicas, PDBs, readiness probes, Multi-AZ deployment, and controlled node draining.
Pod is stuck in Pending — what would you check?
A Pod stuck in Pending means the kube-scheduler cannot find a node that meets all the pod's constraints, or volume mounting / admission controllers are blocked. My primary tool is immediately 'kubectl describe pod <pod-name>'.
Key Architectural Takeaway: Always run 'kubectl describe pod' and read the Scheduler Events. Differentiate resource request starvation (fit calculation) from affinity/taint rules and AZ-locked EBS volume binding.
Pod is in CrashLoopBackOff — how would you troubleshoot?
CrashLoopBackOff means the container started, crashed, and Kubernetes is backing off before restarting it. My troubleshooting sequence always follows: Exit Code analysis -> Previous container logs -> Liveness probe evaluation.
Key Architectural Takeaway: Look at the Exit Code first (137 = OOM, 1 = App crash). Use 'kubectl logs --previous' to capture the crash stack trace. Verify startupProbe isn't killing slow-starting applications.
Pod is running but Service isn't accessible — what's your approach?
When a Pod is Running but its Service isn't accessible, I isolate the failure across 5 discrete layers: Service Selector/Endpoints -> Port mapping -> Readiness probes -> Cluster DNS -> NetworkPolicies.
Key Architectural Takeaway: Always check 'kubectl get endpoints <svc>' first. If endpoints are empty, it's either a label selector mismatch or a failing readiness probe. Then test port mappings and NetworkPolicies.
New deployment breaks production — how would you rollback?
When a new deployment causes production degradation, the priority is mean time to recovery (MTTR). The rollback path depends on whether you run imperative deployments or declarative GitOps.
Key Architectural Takeaway: Know your deployment mechanism: In pure K8s, use 'kubectl rollout undo'; in GitOps (ArgoCD), disable auto-sync or 'git revert' to prevent self-heal fighting. Never do destructive DB migrations in single-step releases.
🌐
Explore Related DevOps & Cloud Domains
Cross-train across interconnected systems for senior and staff infrastructure rounds.
Master Kubernetes & Platform Engineering In Production
Preparing for a cloud or platform interview? Subscribe to get our weekly deep-dive scenario plus instant access to the Top 50 Kubernetes Interview Questions & Incident Runbooks PDF.