The short answer: DevOps interviews test whether you can ship changes quickly and safely and keep systems reliable: CI/CD pipelines, deployment strategies, infrastructure as code, containers and Kubernetes, monitoring and incident response, security in the pipeline, and cloud cost. Many interviewers measure delivery with DORA’s metrics, so know them. Expect hands-on or scenario questions (“A deploy just broke production. What do you do?”) and sometimes a live troubleshooting exercise. Below are the common questions with short answers.
CI/CD
“Walk me through a CI/CD pipeline you built.”
Commit triggers build and unit tests, static analysis and security scans, artifact creation (for example a container image), deployment to a test environment with integration tests, then promotion to production with approvals or automated checks, and monitoring after release. Mention the tools and what you improved.
“What’s the difference between continuous delivery and continuous deployment?”
Continuous delivery keeps every change releasable, with a manual decision to deploy; continuous deployment releases every change that passes the pipeline automatically.
“Compare blue-green, canary and rolling deployments.”
Blue-green switches traffic between two full environments for instant rollback; canary sends a small share of traffic to the new version and expands if metrics are healthy; rolling replaces instances gradually. Feature flags separate deploying code from releasing features.
DORA metrics
“How do you measure software delivery performance?”
DORA, Google Cloud’s research program, measures throughput with change lead time (from commit to production), deployment frequency and failed deployment recovery time, and instability with change fail rate and deployment rework rate. Say which ones you tracked and what you changed to improve them.
Sample answer: “Our change lead time was about nine days because every release waited for a weekly manual test cycle. We automated the regression suite, added canary releases with automatic rollback, and moved to daily deployments. Within a quarter, lead time dropped to under a day and our change fail rate went from about 18% to 7%, because releases were smaller.”
Infrastructure as code and containers
- “Why use infrastructure as code?” Repeatable, reviewable, version-controlled environments; less drift; easier recovery. Tools include Terraform and cloud-native options.
- “How do you manage Terraform state safely?” Remote state with locking, restricted access, separate state per environment, and plan review before apply.
- “What’s the difference between a container and a virtual machine?” Containers share the host kernel and are lighter; VMs virtualize hardware with their own OS.
- “A pod keeps restarting. How do you debug it?” Check events and logs (
kubectl describe,kubectl logs --previous), exit codes, resource limits (out-of-memory kills), probes and configuration. See our Kubernetes interview questions.
Monitoring and incidents
“What’s the difference between monitoring and observability?”
Monitoring tracks known signals and alerts on them; observability (metrics, logs and traces) lets you ask new questions to understand unknown failures.
“A deploy just broke production. What do you do?”
Restore service first: roll back or turn off the feature flag, communicate in the incident channel, confirm recovery with metrics, then investigate the root cause and hold a blameless review with follow-up actions such as a missing test or alert. See our site reliability engineer interview questions.
Security and cost
- “How do you add security to the pipeline?” Dependency and image scanning, secret scanning, least-privilege pipeline credentials, signed artifacts and policy checks. See our security engineer interview questions.
- “How do you reduce cloud costs?” Right-size resources, autoscale, schedule non-production environments off-hours, use committed-use or spot capacity where appropriate, and tag resources to show teams what they spend. See our FinOps analyst interview questions.
Behavioral questions
- Tell me about an outage you handled. See solving a complex problem.
- Tell me about convincing developers to change how they work. See handling conflict.
Frequently asked questions
What questions are asked in a DevOps interview?
What are the DORA metrics?
What's the difference between continuous delivery and continuous deployment?
How do I answer "A deploy broke production"?
Which tools should a DevOps engineer know?
Sources
- DORA, DORA’s software delivery performance metrics. Read on 1 October 2026.
