Command Palette

Search for a command to run...

Hectal

Guide G11 · DevOps path

DevOps Interviews and Career: Turning Skills Into a Job

How DevOps/SRE/platform interviews are run, a framework for troubleshooting questions, infrastructure design rounds, live-debugging scenarios to practise, building a portfolio from the 12 projects, and growing from junior to senior.

Beginner 40 min

Start here

The mental model

Interviewers for DevOps roles aren't checking whether you've memorised flags. They're checking three things: can you REASON through an unfamiliar failure calmly and systematically, do you understand WHY each piece of the stack exists and how they connect, and have you actually BUILT and OPERATED things. Every course here was built around those three, and this guide shows how to demonstrate them.

Go deeper

How it works inside

01What the rounds look like

Typical loop: a recruiter screen; a technical screen (Linux, networking, containers, a cloud); a LIVE TROUBLESHOOTING round ('this service can't reach the database, here's a terminal', or a verbal scenario); an INFRASTRUCTURE DESIGN round ('design CI/CD and hosting for this microservices app', 'how would you make this highly available'); sometimes coding/scripting (Bash or Python: parse a log, call an API, automate a task); and a BEHAVIOURAL round (incidents you handled, disagreements, ownership). SRE roles add more depth in reliability, SLOs, and incident response; platform roles add developer experience and API design for internal tools.

02A framework for troubleshooting questions

1) CLARIFY scope and impact: who's affected, since when, what changed? 2) Form HYPOTHESES by layer: DNS → network path → port → application → dependencies (Networking course, Topic 6.1). 3) Choose the CHEAPEST check that splits the hypotheses, and say which command you'd run and what each outcome would mean. 4) MITIGATE before root-causing if users are affected (roll back, scale, fail over). 5) Explain how you'd PREVENT it (alert, test, runbook). Thinking aloud in this structure matters more than finding the answer first.

A framework for troubleshooting questionsdiagram
Rendering diagram…

03Design rounds

Structure: requirements (traffic, latency, availability target, data, compliance, team size) → high-level architecture (edge, compute, data, async) → delivery pipeline (CI, registry, GitOps, environments) → reliability (multi-AZ, scaling, backups, DR, SLOs) → security (identity, secrets, network, supply chain) → observability (metrics, logs, traces, alerts) → cost → trade-offs and what you'd do first. The Production Architecture guide (G1) and Project 12 are exactly this; practise explaining them in 30 minutes with a diagram.

04Portfolio and résumé

Put three to five of the 12 projects on GitHub with a clear README (problem, architecture diagram, how to run it, what you learned, what you'd improve), the IaC, pipelines, and screenshots or a short demo video. Project 12 alone, done well, carries most interviews. Each project page has a résumé bullet: quantify where you can (build time cut from 12 to 4 minutes, image size from 812 to 241 MB, MTTR from an hour to minutes). Write a short blog post or ADR for one hard decision; it shows judgement.

05Growing from junior to senior

JUNIOR: completes well-defined tasks with the team's tools, learns the stack, follows runbooks. MID: owns services or components end to end, automates toil, handles incidents, improves pipelines. SENIOR: designs systems and platforms, makes trade-offs explicit (ADRs), leads incidents and postmortems, mentors, and measures impact (DORA, SLOs, cost). STAFF/PRINCIPAL: sets technical direction across teams, builds the platform strategy, and multiplies other engineers. Certifications (AWS SAA/SAP, CKA/CKAD/CKS, Terraform Associate) help with screening, especially early on, but projects and operating experience win interviews.

Do it

Hands-on lab

  1. 1

    Practise the 60-second answer

    For each project you put on your résumé, prepare this structure and say it out loud until it takes about a minute.

    project-pitch.mdwhole filemarkdown
    **Context:** ShopLite needed safe, frequent deploys to EKS; deploys were manual helm upgrades from laptops.
    **What I built:** GitHub Actions (OIDC → ECR) + Argo CD with PR-based promotion + Argo Rollouts canaries with Prometheus analysis.
    **Hard part:** canary analysis on low traffic produced inconclusive results → added synthetic k6 traffic during analysis.
    **Result:** deploys went from weekly to several per day; a faulty release was auto-rolled back at 10% traffic within 3 minutes.
    **Next:** preview environments per PR and signed-image verification at admission.

3am practice

Failure drills

Each drill is a real failure mode. Read the scenario and the output, decide what's wrong, then reveal the diagnosis.

Drill #1

Interview scenario: the website is slow

"Users say the site got slow in the last hour. You have kubectl, Grafana, and cloud console access. Walk me through it." The first thing you check shows this:

terminal
$ kubectl -n shop top pods | sort -k2 -h | tail -3; kubectl -n shop get hpa
── what you'll see ──
api-6d8f9c7b5-x4k2p 498m 612Mi
api-6d8f9c7b5-p7r2t 499m 598Mi
api-6d8f9c7b5-q9w7n 500m 605Mi
NAME REFERENCE TARGETS MINPODS MAXPODS REPLICAS
api Deployment/api cpu: 99%/70% 3 3 3

Drill #2

Interview scenario: pods can't reach the database

"After a network change, the orders service logs connection timeouts to RDS. Other services are fine." You run a connectivity test from the pod:

terminal
$ kubectl -n shop exec deploy/orders -- sh -c 'getent hosts shop-db.xxxx.rds.amazonaws.com; nc -vz -w 3 shop-db.xxxx.rds.amazonaws.com 5432'
── what you'll see ──
10.20.21.44 shop-db.xxxx.rds.amazonaws.com
nc: shop-db.xxxx.rds.amazonaws.com (10.20.21.44:5432): Operation timed out

Decide

Certifications worth considering

CertificationProvesWhen it helps most
AWS Solutions Architect – AssociateCore AWS architectureEarly career, cloud-heavy roles
CKA / CKAD / CKSHands-on Kubernetes admin / dev / security (performance-based exams)Kubernetes-focused DevOps and platform roles
HashiCorp Terraform AssociateTerraform fundamentalsIaC-heavy roles, screening
AWS DevOps Engineer – ProfessionalCI/CD, monitoring, automation on AWSMid-level AWS DevOps roles
Linux (LFCS / RHCSA)Hands-on Linux administrationInfrastructure and SRE roles

The bigger picture

Connects to

Prove it

Interview questions

01

Tell me about an incident you handled.

02

What happens when you type a URL into a browser and press Enter?