DevOps engineer behavioural interview questions

As asked

Tell me about the worst on-call shift you have personally had. What broke, what did you do in the moment, and what changed afterwards?

Sample answer outline

Pick a real story with measurable user impact. Cover detection (was it self-detected or customer-reported), the diagnosis path (especially the dead ends), the mitigation, and the postmortem actions that actually landed. Interviewers are listening for ownership without blame, calm under pressure, and process changes that prevent the next one. Avoid 'we just restarted it' answers, dig into the root cause.

Expect these follow-ups

• What did the postmortem find as the contributing factors?
• Did the action items actually ship, or did they slip?
• What would you do differently in the first 10 minutes if it happened again?

As asked

Tell me about the biggest incident you handled that lived in the platform layer: a broken deploy pipeline, an infrastructure change gone wrong, a certificate expiry, or a cluster failure. Walk me through detection, mitigation, and what changed in the platform afterwards.

Sample answer outline

Frame it through a platform lens, where the blast radius is every team that depends on you. Describe the impact across consumers, not just one service. Detection: what alerted you, and whether it was your monitoring or a downstream team that noticed first. Mitigation: the rollback or break-glass procedure, and whether it existed before the incident or had to be improvised. The strong answer ends with platform-level prevention: a guardrail in the pipeline, a pre-deploy check, an expiry alert, automated rollback. Interviewers listen for ownership of shared infrastructure and the discipline to turn one painful event into a control that protects every team.

Expect these follow-ups

• Did a self-service guardrail exist, or did you have to build one after?
• How did you communicate with the many teams affected at once?
• What pipeline or infrastructure check would have caught this earlier?

As asked

Tell me about a production outage you caused or contributed to. What happened, what was the impact, how did you fix it, and what did you change afterwards to make sure it could not happen the same way again?

Sample answer outline

Strong answers are specific and honest, describe the sequence of events clearly, quantify the impact (users affected, duration, revenue), show personal ownership without blame-shifting, explain the immediate mitigation, and most importantly describe systemic improvements not just personal behavior changes. The candidate should demonstrate blameless postmortem thinking and that the improvement addressed the root cause not the symptom.

Expect these follow-ups

• How did you communicate the outage to leadership and customers during the incident?
• If you had to roll back a change quickly, what is the fastest path you have used?

As asked

Describe a time when you strongly disagreed with a technical decision made by an architect or senior engineer on an infrastructure or platform topic, such as a cloud architecture choice, a CI/CD design, or a reliability tradeoff. How did you handle it and what was the outcome?

Sample answer outline

Strong answers show that the candidate raised the concern with data and a concrete alternative proposal, not just an objection. They escalated through the right channels and used written communication to create a record. If they lost the argument, they committed to the decision and monitored for the outcomes they predicted. A great answer shows the candidate was right but also shows they can lose gracefully when the decision was reasonable even if not their preference.

Expect these follow-ups

• How do you decide when to escalate a technical disagreement versus just accepting the decision?
• Have you ever changed your mind during a technical debate after hearing the other side?

As asked

Give me an example of repetitive manual work that was consuming a significant amount of your team's time. How did you identify it, justify the automation investment, build the solution, and measure the improvement?

Sample answer outline

Strong answers quantify the toil (hours per week times number of people), describe the automation built with specific technologies, measure the reduction in time spent after rollout, and show that the automation was reliable enough that it did not just move toil from manual to maintenance. SRE candidates should reference Google's 50% toil cap concept and explain how they prioritized this work against feature requests.

Expect these follow-ups

• What made you confident the automation was safe to leave running unattended?
• Was there any toil you decided not to automate and why?

As asked

Describe how you have onboarded a new engineer onto a complex infrastructure or DevOps codebase. What systems did you introduce first, how did you teach them to safely make changes to production infrastructure, and what did you do to make them productive in the first two weeks?

Sample answer outline

Strong answers describe a structured onboarding plan: starter issues that touch the critical paths, pairing sessions on the most important systems, documentation they were asked to improve as a first contribution, and access setup with least-privilege starting permissions. The candidate should mention feedback loops (asking the new engineer what was confusing) and show that onboarding improved the documentation for future engineers.

Expect these follow-ups

• What was the most surprising thing a new engineer found confusing about your infrastructure?
• How do you balance onboarding time with your own project commitments?

As asked

Tell me about a time when you were asked to deploy a feature or infrastructure change on a timeline you thought was unsafe. How did you handle the pressure and what was the result?

Sample answer outline

Strong answers show the candidate quantified the specific risks (what could go wrong, what the blast radius would be), proposed a safer alternative timeline or a reduced scope that could meet the deadline, and escalated to the right level without being obstructionist. They should demonstrate comfort with saying no with data and show that safety decisions were respected or explain what happened when they were overruled.

Expect these follow-ups

• How do you frame a risk concern to a non-technical stakeholder who is focused on the business deadline?
• Have you ever been overruled and the risk materialized? What happened?

As asked

Describe a situation where your team's deployment frequency was too low or your deployments were too risky. What changes did you make and what did the metrics look like before and after?

Sample answer outline

Strong answers cite specific metrics (deploys per week, change failure rate, MTTR) before and after. The changes should be concrete: automated testing gates, smaller batch sizes, feature flags, automated rollback on alert threshold breach. The candidate should explain how they got buy-in from engineers who were comfortable with the old process and how they measured that the new process was actually safer and not just faster.

Expect these follow-ups

• How did you convince skeptical engineers to deploy more frequently when their intuition said big releases are safer?
• What metric best correlates deployment frequency with risk in your experience?

As asked

Tell me about a time when your work was blocked by a dependency on another team's infrastructure or platform. How did you unblock yourself or resolve the dependency?

Sample answer outline

Strong answers show proactive communication, escalation when appropriate, and creative unblocking strategies (building a temporary workaround, contributing the fix to the other team's repo, or finding an alternative). The candidate should show they understand organizational dynamics and that they did not just wait passively or escalate immediately to management without trying direct collaboration first.

Expect these follow-ups

• How do you maintain momentum on your own work while waiting for an external dependency?
• What would you do differently if you encountered the same situation again?

As asked

Describe a time you mentored a junior engineer who was new to infrastructure or DevOps work. What did you focus on first, how did you give feedback on their pull requests, and how did you know when they were ready to work independently?

Sample answer outline

Strong answers show intentionality: starting with concepts like immutability and idempotency before diving into specific tools, giving PR feedback that explains the why not just the what, and pairing on the scariest tasks rather than saving them for later. The signal for independence is not time elapsed but demonstrated ability to reason through a novel problem without prompting. Good candidates also mention the two-way learning: the junior often spotted assumptions the senior had stopped questioning.

Expect these follow-ups

• How do you give critical feedback on a junior's pull request without discouraging them?
• What is the most important infrastructure concept you make sure every junior on your team understands early?

As asked

Tell me about a time you discovered a security vulnerability in a running production system. How did you assess the risk, handle disclosure, and remediate it while keeping the service running?

Sample answer outline

Strong answers show a clear risk assessment process (exploitability, blast radius, data exposure), responsible handling (not publicizing before patching), a prioritized remediation plan, and communication with legal/security teams if customer data was at risk. The candidate should demonstrate they understand the difference between a vulnerability and an active exploit and that speed of patching depends on severity.

Expect these follow-ups

• How did you verify the vulnerability was actually exploitable and not just a theoretical risk?
• What monitoring did you add after remediating to detect similar issues going forward?

As asked

Tell me about a time you identified and led a significant cloud cost reduction initiative. How did you find the savings opportunity, prioritize it, and execute without disrupting running services?

Sample answer outline

Strong answers start with the discovery (Cost Explorer analysis, right-sizing recommendations, idle resource audit), quantify the savings opportunity, describe the execution plan (phased approach, testing changes in non-prod first), and give the final number (e.g. $20k/month saved). They should mention how they verified services continued to function correctly and how they communicated the savings to stakeholders.

Expect these follow-ups

• How did you decide which cost optimizations were worth the operational risk of changing running infrastructure?
• What monitoring did you add to make sure the cost reduction held over the following months?

As asked

Your team is exhausted from noisy alerts and frequent overnight pages. Tell me about how you have addressed on-call burnout, either on your own team or on one you have joined.

Sample answer outline

Strong answers quantify the problem (pages per night, alert-to-action ratio, how many were actionable), describe a process of categorizing alerts by type and removing non-actionable ones, setting alert thresholds with intent, and treating runbook gaps as technical debt. The candidate should mention alert postmortems as a regular practice and show measurable improvement in pages per week. Mention the SRE concept of alert quality over alert quantity.

Expect these follow-ups

• How do you decide whether an alert is genuinely actionable or just informational noise?
• How do you get engineers to agree to delete an alert they created, even if they still think it might be useful someday?

As asked

Describe how you have contributed to or built a blameless postmortem culture at a company or team. What resistance did you encounter and how did you overcome it?

Sample answer outline

Strong answers explain the principles of blameless postmortems: the system failed, not the individual, and the goal is systemic improvement not punishment. They describe facilitating postmortems, asking 5-whys questions, framing mistakes as learning opportunities, tracking action items to completion, and sharing postmortems across the organization. Resistance usually comes from leadership who want someone to blame; the candidate should show how they reframed this as a risk reduction strategy.

Expect these follow-ups

• How do you handle a postmortem where someone did make a genuine error of judgment?
• What makes a postmortem action item actually get completed versus sitting in a backlog forever?