DevOps and cloud engineer interview questions
DevOps and cloud interviews test whether you can build, ship and run systems reliably: Linux and networking basics, pipelines, containers, a cloud provider, infrastructure as code, and how you behave when production breaks.
The questions below are grouped by level, with what a strong answer should cover.
What interviewers assess
- Linux, networking and scripting fundamentals
- CI/CD pipelines, containers and orchestration
- Cloud services, infrastructure as code, security and cost
- Monitoring, incident response and learning from failures
Basic questions
What happens when you type a URL into a browser?
What a good answer covers: DNS lookup, TCP and TLS handshakes, the HTTP request, load balancer, server and response. Shows networking fundamentals in one question.
What is the difference between a container and a virtual machine?
What a good answer covers: Containers share the host kernel and isolate processes; VMs virtualise hardware with their own OS. Trade-offs in startup time, density and isolation.
Explain what a CI/CD pipeline does.
What a good answer covers: Build, test and scan on every change, then deploy automatically or with approval. Mention fast feedback, artefacts and rollbacks.
How do you check why a Linux server is slow?
What a good answer covers: top or htop, free, df, iostat and logs; check CPU, memory, disk and network in turn rather than guessing.
What is infrastructure as code?
What a good answer covers: Defining infrastructure in version-controlled files (Terraform, CloudFormation, Bicep) so it is repeatable, reviewable and consistent across environments.
Intermediate questions
How do you make a Docker image smaller and more secure?
What a good answer covers: Multi-stage builds, slim base images, fewer layers, .dockerignore, a non-root user, pinned versions and image scanning.
Explain Kubernetes Deployments, Services and Ingress.
What a good answer covers: A Deployment manages replicas and rollouts; a Service gives a stable address to pods; Ingress routes external HTTP traffic. Mention health probes.
How do you manage secrets in pipelines and applications?
What a good answer covers: A secrets manager or vault, never committing secrets, short-lived credentials, least privilege and rotation.
How would you set up deployments with no downtime?
What a good answer covers: Rolling, blue-green or canary deployments, health checks, backwards-compatible database migrations and automatic rollback.
What do you monitor for a web service?
What a good answer covers: The four golden signals (latency, traffic, errors, saturation), logs and traces, alerts tied to user impact, and dashboards.
Advanced questions
Production is down. Walk me through what you do.
What a good answer covers: Acknowledge and communicate, find and limit the impact, roll back if a recent change is suspected, then diagnose, fix and run a blameless post-mortem with follow-up actions.
Design a highly available setup for a web app on AWS, Azure or GCP.
What a good answer covers: Multiple availability zones, load balancer, auto-scaling, a managed database with replicas, backups, health checks and infrastructure as code. Discuss cost.
How do you manage Terraform state for a team?
What a good answer covers: Remote state with locking, separate state per environment, modules, code review and plan output in CI.
How would you reduce a cloud bill that keeps growing?
What a good answer covers: Find the biggest costs first, right-size resources, use auto-scaling, reserved or spot capacity, storage lifecycle rules and tagging for ownership.
Define SLIs, SLOs and error budgets.
What a good answer covers: An indicator you measure, a target for it, and the allowed failure budget that balances reliability against releasing faster.
Common mistakes
- Naming tools without explaining the problem they solved for you.
- Describing an incident without mentioning communication or a post-mortem.
- Ignoring security and cost in a design answer.
- Claiming Kubernetes experience from tutorials only and struggling with follow-ups.
How to prepare
- Deploy a small app end to end: Dockerfile, a CI pipeline, cloud hosting and monitoring.
- Practise Linux troubleshooting commands until you can use them without looking them up.
- Prepare one incident story with what happened, what you did and what changed afterwards.
- Revise the core services of the cloud provider named in the job.
Reading questions only goes so far: the real test is answering them aloud, under follow-up questions. Career Scan builds a mock interview from your own resume and the job, then shows exactly which answers need work.
Frequently asked questions
Which cloud provider should I learn for interviews?
The one named in the job description. AWS is the most common, but Azure and GCP concepts map closely to it.
Are certifications important?
They can help you get shortlisted, but interviews test practical experience. Be ready to show something you built.
Do DevOps interviews include coding?
Usually scripting (Bash or Python) and sometimes writing a pipeline or Terraform configuration.