Interview prep · 🔁 DevOps / SRE Engineer
DevOps / SRE Engineer interview questionsand answers (30, by level).
The questions Indian employers actually ask for this role — fresher, mid-level and senior — with model answers written by engineers who sit on the other side of the table. Use with the role path and the 30-day lab plan.
Fresher / L1
Fundamentals they test on day one
1What is CI/CD?
Continuous Integration: build/test every change; Continuous Delivery/Deployment: automatically ship to environments with gates.
2Git: branch, merge, rebase — difference?
Branch isolates work; merge combines with a merge commit; rebase replays commits on a new base for linear history.
3What is Docker and a Dockerfile?
Containers package app + dependencies; Dockerfile defines the image layers (FROM, RUN, COPY, CMD).
4What is infrastructure as code?
Defining infra in versioned code (Terraform, Ansible) for repeatability and review.
5Explain a Kubernetes Pod, Deployment and Service.
Pod: one or more containers; Deployment: desired replicas with rolling updates; Service: stable virtual IP/DNS to reach pods.
6What are environment variables and secrets in pipelines?
Config injected at runtime; secrets stored in a vault/CI secret store, never in code.
7What is monitoring vs logging vs tracing?
Metrics (numbers over time), logs (events), traces (request paths across services) — the three pillars.
8What is a reverse proxy?
A server in front of apps handling TLS, routing, caching (NGINX, Traefik).
9Explain YAML pitfalls.
Indentation with spaces, quoting strings like yes/no, tabs forbidden — our YAML tool validates.
10What does “immutable infrastructure” mean?
Replace servers/images rather than patching in place; reduces drift.
2–5 years / L2
What a working engineer must know
1Design a CI/CD pipeline for a microservice.
Lint/test → build image → scan (Trivy) → push to registry → deploy to staging via GitOps → integration tests → promote to prod with approval/canary → rollback plan.
2What are SLIs, SLOs, error budgets?
SLI: measured indicator (latency, availability); SLO: target; error budget: allowed failure that governs release pace.
3Kubernetes: requests vs limits; what happens when exceeded?
Requests schedule; limits cap; CPU throttles, memory OOM-kills. Set both from measurements.
4How do you debug a CrashLoopBackOff?
kubectl describe/logs --previous, check probes, env/config, image tag, resources, dependencies.
5GitOps with Argo CD — benefits and gotchas?
Git as source of truth, drift detection, audit; gotchas: secrets handling (SOPS/External Secrets), sync waves, app-of-apps structure.
6Prometheus alert for high error rate — write it.
rate(http_requests_total{status=~"5.."}[5m]) / rate(http_requests_total[5m]) > 0.05 for 10m; label severity; route via Alertmanager. Our alert-rules generator helps.
7Blue/green on Kubernetes without downtime?
Two Deployments + Service selector switch, or Argo Rollouts; readiness probes; DB migrations backward-compatible.
8Terraform module design principles?
Small reusable modules, inputs/outputs, versioned via Git tags, no hardcoded env, tested with plan in CI.
9Secrets management for Kubernetes?
External Secrets Operator with Vault/cloud secret manager, sealed-secrets, RBAC on namespaces; never plain Secrets in Git.
10On-call: how do you reduce pages?
Alert on symptoms not causes, tune thresholds, runbooks, auto-remediation, fix top recurring issues, postmortems.
Senior / L3
Design, judgement and leadership
1Build an internal developer platform for 30 teams.
Golden paths (templates), self-service environments, GitOps, standardised observability, policy as code, cost visibility, documentation; measure DORA metrics.
2Describe a chaos engineering programme.
Hypotheses, controlled experiments (pod kill, latency, AZ loss), game days, blast radius limits, learnings into backlog.
3Multi-cluster Kubernetes strategy?
Clusters per region/env, central GitOps, shared identity, service mesh only when needed, cluster API or managed services; avoid one giant cluster.
4How do you secure the software supply chain?
Signed commits/images, SBOM, dependency scanning, pinned actions, least-privilege CI with OIDC, admission controls verifying signatures.
5Database schema changes in CD?
Expand/contract pattern, backward-compatible migrations, feature flags, migration jobs with locks/timeouts, rollback strategy.
6Cost control for Kubernetes?
Right-size requests (VPA), autoscaling (HPA/Karpenter), spot nodes for stateless, namespace quotas, cost allocation (OpenCost), idle cleanup.
7How do you run blameless postmortems that change things?
Timeline, contributing factors, action items with owners/dates tracked like features, share widely.
8Observability at scale: cardinality and cost?
Limit labels, aggregate, sampling for traces, tiered retention, Loki/Mimir or VictoriaMetrics for cost.
9Migrating Jenkins to GitLab CI/GitHub Actions — approach?
Inventory pipelines, standard templates, migrate new services first, parallel run, decommission; see our comparison pages.
10What DORA metrics would you report and target?
Deployment frequency, lead time, change failure rate, MTTR; targets per team maturity, trend over time.
How to answer
Four habits that pass interviews
- Say the structure first“I’d check three things: …” — interviewers reward a method more than a lucky answer.
- Numbers and namesMention real tools, versions and figures from your lab; vague answers lose.
- Admit gaps cleanly“I haven’t run X in production; here is how I’d approach it” beats bluffing.
- Ask about the environmentScale, on-call, tooling — it shows you think like an operator.
Ready to be hired, or ready to hire?
Engineers: join our vetted network for projects and placements. Companies: we shortlist devops / sre engineers within 48 hours.