Interview Prep Guide

Kubernetes Interview Questions and Answers for DevOps and Platform Roles

Prepare for Kubernetes interviews with structured questions on pods, deployments, services, scaling, reliability, and cluster operations.

Basic Kubernetes Interview Questions

What problem does Kubernetes solve beyond using Docker alone?
Kubernetes helps manage containerized workloads at scale by handling deployment, service discovery, self-healing, rollout control, and scaling across multiple machines.
What are pods, deployments, and services in Kubernetes?
Pods are the basic running units, deployments manage desired replica state and rollouts, and services provide stable network access to sets of pods.

What are liveness, readiness, and startup probes used for?
They help Kubernetes understand whether a workload has started correctly, is healthy, and is actually ready to receive traffic.
How do ConfigMaps and Secrets fit into Kubernetes application design?
They separate configuration from the application image so teams can manage environment-specific settings and sensitive values more safely and flexibly.

How do you think about scaling, rollouts, and reliability in Kubernetes?
Think in terms of traffic patterns, safe deployment strategy, resource constraints, failure recovery, and observability rather than treating autoscaling as a magic switch.

How would you debug a service in Kubernetes where pods keep restarting and user traffic is failing intermittently?
Inspect pod events, logs, probe behavior, resource pressure, configuration changes, and service-to-pod routing before assuming the application code is the only cause.

How do readiness, liveness, and startup probes differ?
Readiness controls whether a Pod receives traffic, liveness decides when a stuck container should restart, and startup delays the other probes until a slow application has initialized.
What is the difference between resource requests and limits?
Requests guide scheduling and reserve expected capacity; limits constrain runtime consumption according to the resource and container runtime behavior.
How does a Kubernetes Service find changing Pods?
A Service selects eligible Pods and exposes stable discovery while EndpointSlices track the current ready network endpoints.
How would you investigate a CrashLoopBackOff?
Start with current and previous container logs, exit reason, events, probe failures, configuration, secrets, permissions, and resource termination details.
How do rolling updates protect availability?
A Deployment gradually replaces old Pods while surge, unavailable, readiness, and progress settings control capacity and rollout safety.

What is the difference between a Deployment and a StatefulSet?
A Deployment manages interchangeable Pods, while a StatefulSet provides stable identities, ordered behavior, and persistent-storage association for stateful workloads.
How do ConfigMaps and Secrets differ, and what security limits remain?
Both inject configuration, but Secrets are intended for sensitive values and receive specialized handling; they still require encryption, access control, and careful delivery.
What happens when a Pod is evicted?
Kubernetes terminates the Pod due to node pressure or policy, and a controller may create a replacement elsewhere if the workload declares a desired replica count.

Write or explain a deployment/service configuration for a small app and justify the probe, scaling, and service choices
Strong answers usually explain what each resource is doing, why the workload is exposed in a certain way, and how the deployment could be made safer or more observable in production.