What does a DevOps engineer do?
A DevOps engineer builds and runs the systems that move code from a developer's laptop to production and keep it healthy there. The work covers build and deploy pipelines, infrastructure written as code, monitoring and alerts, and incident response. Most teams expect you to write code and to be on call for what you ship.
The Bureau of Labor Statistics has no separate profile for DevOps engineers. The nearest is software developers, who "design computer applications or programs." BLS gives a median annual wage for software developers of $135,980 in May 2025. It projects the wider group to grow 10 percent from 2025 to 2035, "much faster than the average for all occupations."
O*NET OnLine lists "DevOps Engineer (Development Operations Engineer)" as a reported job title under Software Developers (15-1252.00). Its tasks include "Coordinate installation of software system" and "Monitor functioning of equipment to ensure system operates in conformance with specifications." The operations side overlaps with network and computer systems administrators, who BLS says "install, configure, and maintain organizations' computer networks and systems."
AWS's page What is DevOps? defines it as "the combination of cultural philosophies, practices, and tools that increases an organization's ability to deliver applications and services at high velocity." It names practices such as continuous integration, continuous delivery, infrastructure as code, and monitoring and logging. Those practices are the backbone of most technical rounds.
In Canada, Job Bank files DevOps engineers under NOC 21231, software engineers and designers. The page, modified September 24, 2026, lists a median wage of $56.49 an hour and says the occupation "usually requires a university degree."
How were these DevOps engineer questions chosen?
They come from the duties in the O*NET and BLS profiles and from the practices AWS names, each matched to the question it tends to produce. They are drawn from the job's published duties, not from a survey of real interviews. Nothing here claims to be the most asked list or to match what a named company asks.
| Duty or practice (O*NET, BLS, AWS) | Question it tends to produce | What the interviewer is checking |
|---|---|---|
| Coordinate installation of software system | "Walk me through a deploy pipeline you built" | Each stage, why it exists, and how a bad change gets stopped |
| Provision infrastructure as code | "How do you change production infrastructure safely?" | Version control, review, a plan step and a way back |
| Monitor functioning to meet specifications | "How would you know this service is healthy?" | Measures tied to what users feel, and alerts people act on |
| Diagnose and resolve problems from monitoring alerts | "You're paged at 2 a.m. What do you do first?" | Calm order: limit harm, tell people, then find the cause |
| Perform backups and disaster recovery | "How would you test a restore?" | Whether you have restored a backup, beyond taking one |
| Work across development and operations | "Tell me about a time a team pushed back on your change" | How you explain risk and agree on a plan |
Which behavioral questions should you expect?
Behavioral questions ask about past work, so answer each with one real event in STAR order: situation, task, action, result. Interviewers listen for your own part in an incident or a migration, how you kept others informed, and what you changed afterwards. Keep each answer near two minutes and name tools only where they matter.
The guide to the STAR method covers the structure. These four questions follow from the duties in the table.
1. Tell me about an outage you were part of.
What they listen for: what you did in the first minutes, who you told, how the cause was found, and what changed after. Google's SRE book chapter on postmortem culture says a blameless postmortem "must focus on identifying the contributing causes of the incident" without "indicting any individual or team," so avoid blaming a colleague.
Made-up example answer. Situation: a configuration change I merged on a Friday caused one service to reject most logins. Task: I was on call and the change was mine. Action: I rolled the change back within minutes, posted in the incident channel what I had done and what I knew, then compared the old and new config to find the bad value. Result: logins recovered after the rollback, and in the review I proposed a test that loads each config in a staging copy before merge, which the team adopted.
2. Tell me about a manual process you automated.
What they listen for: why it was worth automating, how you made the script safe to rerun, and how you measured the time or errors saved. Say who uses it now.
3. Describe a time a developer team resisted a change you wanted.
What they listen for: the reason behind their pushback, the evidence you brought, and the agreement you reached. The page on influence without authority fits this question well.
4. Tell me about a mistake you made in production.
What they listen for: ownership, a quick fix, and a lasting guard against repeats. The guide to tell me about a mistake you made has a structure that works here.
What role-specific DevOps engineer questions come up?
Role-specific questions test the work itself: pipelines, infrastructure as code, monitoring and reliability targets. Interviewers check whether you explain why each step exists, what happens when it fails, and how you would roll back. Naming a tool is less useful than saying what it protects you from and what it costs.
5. Walk me through a CI/CD pipeline you would build for a web service.
What they listen for: build, automated tests, a security or dependency check, a staged rollout and a fast rollback. AWS describes continuous delivery as code changes that are "automatically built, tested, and prepared" for release to production.
Made-up example answer: "Every merge to main triggers a build and the unit tests. If those pass, the pipeline builds one image and runs integration tests against a staging copy. Then it deploys to a small slice of production and watches error rate and latency for ten minutes. If either crosses the limit we set, it rolls back on its own and posts in the team channel. Otherwise it continues to the rest of the fleet."
6. What is infrastructure as code, and how do you review a change to it?
What they listen for: a clear definition and a safe review habit. AWS says infrastructure as code means infrastructure "is provisioned and managed using code and software development techniques, such as version control and continuous integration." Mention reviewing the planned diff before applying it.
7. What is the difference between an SLI, an SLO and an SLA?
What they listen for: exact definitions and an example. The SRE book chapter on service level objectives defines an SLI as "a carefully defined quantitative measure of some aspect of the level of service," an SLO as "a target value or range of values for a service level that is measured by an SLI," and an SLA as a contract with users that includes consequences of meeting or missing the SLOs.
8. How would you measure whether your team's delivery is improving?
What they listen for: a small set of measures and a reason for each. DORA's guide to software delivery performance metrics, last updated January 5, 2026, lists change lead time, deployment frequency, failed deployment recovery time, change fail rate and deployment rework rate.
9. When should a team stop shipping features and work on reliability?
What they listen for: a rule agreed in advance. The SRE book chapter on embracing risk describes an error budget based on the SLO, and says that if violations use up the budget, "releases are temporarily halted."
Which situational questions test judgment?
Situational questions describe a problem you have not faced and ask what you would do first. Interviewers want an order of steps: stop the harm, tell the right people, find the cause, then prevent it next time. Ask one or two questions before you answer, because jumping straight to a fix can sound like guessing.
The guide to situational interview questions explains how these differ from behavioral ones.
10. Latency doubles right after a deploy. What do you do?
What they listen for: rolling back first if the timing points to the deploy, then checking dashboards for what changed, then telling the team.
Made-up example answer: "If the jump lines up with the deploy, I'd roll back first, since that's the fastest way to help users. I'd post in the incident channel that latency doubled, that I rolled back, and that I'm checking whether it recovered. If it did, I'd compare the two versions and the deploy's config. If it didn't, the deploy probably wasn't the cause, so I'd look at the database, the network and any traffic change."
11. A secret key is found in a public repository.
What they listen for: revoking and rotating the key at once, checking logs for use of it, removing it from history, and telling security. Deleting the commit alone does not make the key safe again.
12. Leadership wants to move every service to a new platform this quarter.
What they listen for: a pilot with one low-risk service, a clear way back, and an honest estimate shared early. The page on competing priorities helps with the scoping part.
What follow-up questions tend to come next?
Follow-ups test whether your first answer holds up. Expect the interviewer to break one part of your design, ask how you would know it broke, or ask what you would do with half the time. Prepare by asking yourself what fails first in each system you describe, and how you would find out.
- "What happens if the deploy tool itself is down?"
- "How would you roll back a database migration?"
- "Which alert in your last job woke people up for nothing?"
- "How did you know the fix worked?"
- "What part of that incident were you personally responsible for?"
How should you prepare for a DevOps engineer interview?
Write four or five STAR stories: an outage, an automation, a migration, a disagreement and a mistake. Practise drawing one pipeline and one service's monitoring out loud with a timer. Read the posting again and match each tool or duty it lists to a story or a project you can explain in plain words.
The free interview question generator lets you paste the job and pick the round, and it returns likely questions, why they come up, and what the interviewer listens for. DevOps loops often overlap with other engineering roles, so compare this page with the questions for a software engineer, a data engineer or an engineering manager.
Check that your resume shows pipelines, migrations or incidents with results you can talk through. The guide to resume bullet points shows how to write them. Prept's Mock room lets you answer out loud on a clock, scores each answer out of 10 and checks it against your resume, and it is included in the free plan.