DevOps interviews are not like other interviews.
You don't just get asked "what is a pod" or "what does S3 stand for." You get asked to think out loud. Walk me through how you'd debug a service that's returning 503s in production. What would you check first if your Kubernetes nodes are NotReady at 2am? How do you handle a Terraform state conflict when two engineers applied at the same time?
Most interview prep resources aren't built for that. They're flashcard decks. Multiple choice questions. Definition lists. None of that prepares you for the moment an interviewer says "okay, now walk me through it."
That's the gap I was trying to close.

What I Built

A few months ago I started building a GitHub repo called devops-cloud-interview-scenarios.
The idea was simple: scenario-based questions, written the way interviewers actually ask them, with plain-language answers and follow-up questions included. Not definitions. Not theory. Actual situations you'd face in a real role.
The repo now has 770+ scenarios across 11 domains — Kubernetes, AWS, CI/CD, Terraform, Docker, Linux/SRE, Observability, Networking, Security, Git, and General DevOps. Every scenario is tagged by difficulty: L1 for fresher-level, L2 for mid-level, L3 for senior.

A few examples of what a scenario looks like:

Situation: Your team just migrated a monolith to microservices on Kubernetes. Two weeks after go-live, intermittent latency spikes appear with no obvious pattern. How do you approach this?

Situation: A Terraform apply is failing with "Error acquiring the state lock." Another engineer says they didn't run anything. What do you do?

Situation: Your CI pipeline is passing but the production deployment is failing silently — the pods start, pass health checks, then crash ten minutes later. Walk me through your debugging approach.

That's the format. Situation, what the interviewer is really testing, a structured answer, and follow-up questions that a good interviewer will actually ask next.

github.com/Techikrish/devops-cloud-interview-scenarios

The Repo Needs Your Scenarios

Here's something I want to be upfront about.
I built this from my own research and learning. But the best scenarios in this kind of resource come from people who've actually lived through these situations who've been on-call when things broke, who've sat in post-mortems, who've hit edge cases that no documentation prepared them for.
If that's you, the repo needs what you know.
The domains where gaps are most visible right now are Observability, Networking, and Security. Real Prometheus alerts that turned out to be something completely unexpected. Networking failures that took hours to trace. Security incidents that taught you something the certifications never covered.
Contributing is straightforward. Fork the repo, go to the domain folder that matches your experience, and add your scenario in the standard format — situation, what the interviewer is testing, a plain-language answer, follow-up questions, and an L1/L2/L3 tag. Open a PR and the duplicate-check GitHub Action runs automatically. Full details in CONTRIBUTING.md.
You don't need to write fifty scenarios. Five genuinely good L2 questions from someone who's lived through them are more valuable than fifty generic ones. If you've had an incident that taught you something real, that's a scenario worth adding.

Why I Built a Website on Top of It

The repo works well on a laptop. It's not great on a phone.
And phones are where most people actually do interview prep on a commute, during lunch, the night before a call. The markdown format that looks clean on GitHub becomes a scroll-fest on mobile. There's no search. No filtering by difficulty. No way to jump straight to Kubernetes L2 scenarios without digging through folders.
So I built a proper web interface on top of the same content.

techikrish.github.io/devops-cloud-interview-scenarios

Search across everything. Type "crashloopbackoff" or "terraform state lock" or "prometheus scrape" and it surfaces the relevant scenarios instantly — across all 770+ questions. No folder navigation, no scrolling past irrelevant content.
Difficulty filtering. One click to switch between All, L1, L2, and L3. If you have an interview next week for a senior role, you filter to L3 and work through just those. If you're early in your career and want to build confidence first, start at L1.
Domain picker. Select any of the eleven domains from a clean UI. Pick Terraform and see only Terraform scenarios. Pick Observability and see only Observability scenarios.
Fast, no friction. It's a static site on GitHub Pages. No login, no account, nothing to install. Open it and start practicing.
The difference between a resource you save and one you actually use is usability. The site is built for the second kind.

The Android App Is Coming

The Android app is currently under review on the Google Play Console.
Same content, same search and filter, but native on Android and completely offline. No WiFi needed, no data, no load times. Just open it and practice on the metro, in a waiting room, anywhere.
I'll post an update here on DEV the moment it clears review.

Find Everything

| 🌐 Website | techikrish.github.io/devops-cloud-interview-scenarios |
| ⭐ GitHub Repo | github.com/Techikrish/devops-cloud-interview-scenarios |

Android App Under review on Google Play — coming soon

The goal is simple: build the interview prep resource
One built around real situations, not definitions. One that gets better as more engineers who've actually worked in these environments contribute to it.
If this is useful to you star the repo, share it with someone who has an interview coming up, or contribute a scenario from your own experience. That's how this gets better for everyone.