How to Prepare for an SRE (Site Reliability Engineer) Interview
An SRE interview tests whether you can keep a running service reliable: write practical code, reason about Linux and networks, investigate a failure step by step, set reliability targets, design infrastructure, and lead an incident. Most loops combine those into a screen, one or two coding rounds, a Linux or troubleshooting round, a design round, and a behavioral round. The exact mix changes by company, team, and level, so ask the recruiter which rounds you will face.
This answer explains what each round tests, how SRE preparation differs from DevOps preparation, and a four-week plan.
The Rounds and What They Test
| Round | What you do | What it tests |
|---|---|---|
| Coding | Parse a log, build a small tool, or solve an easy or medium algorithm problem | Practical automation code |
| Linux and networking | Explain processes, memory, file systems, DNS, TCP, and TLS | How systems behave, not only which command to run |
| Troubleshooting | Investigate a described failure, such as a website that is down, one step at a time | Method, evidence, and explanation under time pressure |
| Reliability practice | Discuss SLIs, SLOs, error budgets, alerting, and incident severity | How you decide what "reliable enough" means |
| Infrastructure design | Design a system such as a metrics pipeline, with estimates and failure handling | Concrete capacity and operational thinking |
| Behavioral and project review | Describe incidents you handled, conflicts, and past projects in detail | Ownership, judgment, and communication |
Candidate reports show that behavioral weakness can end a loop even after strong technical rounds, so treat it as a real round.
How SRE Differs From DevOps Preparation
The two roles share most of their rounds. The difference is emphasis. SRE loops ask more about measured reliability. An SLI (service level indicator) is a measurement, such as the share of checkouts that succeed within 1 second. An SLO (service level objective) is the target for it, such as 99.9 percent over 30 days, and the error budget is the failure the target allows. At 900,000 checkouts in 30 days, a 99.9 percent SLO allows 900 bad checkouts.
Expect questions such as "A service meets its availability SLO, yet users say it feels slow. What do you measure first?" and "How do you calculate burn rate, and why does it matter?" DevOps loops more often ask about pipelines, Terraform, and Kubernetes in daily use.
How Answers Change With Level
There is no published standard, but a useful guide is the same question answered at three levels: "Deployments sometimes break checkout. What would you do?"
- Mid-level: reverse the bad change, find the cause, and add a test or check.
- Senior: also limit harm with canary releases, set rollback rules, and lead the incident.
- Staff: also change the release standard for many teams and reach agreement with leaders on acceptable risk.
A Four-Week Plan
- Week 1: Linux, networking, and a troubleshooting method. Learn to read
top,free,vmstat,df,ss, anddig, and practice a method: confirm, scope, mitigate, form a hypothesis, check one thing at a time, and explain each step. - Week 2: coding. Write a log parser, a grep style tool, and a
tail -nin Python from a blank file, and solve easy and medium array and hash map problems. - Week 3: reliability and incidents. Calculate error budgets and burn rates, design alerts that page only for user-visible problems, and read two or three public postmortems.
- Week 4: design and stories. Practice two infrastructure designs with estimates, and prepare five incident or project stories with numbers and what changed afterward.
Grokking DevOps and SRE Interviews covers these rounds for DevOps, SRE, platform, and production engineer roles, with troubleshooting drills, SLO and burn rate lessons, infrastructure design questions, and two full practice interviews. For behavioral preparation in general, see Grokking Modern Behavioral Interview.
Key Points
- SRE loops test coding, Linux and networking, troubleshooting, reliability practice, design, and behavioral skills.
- SLIs, SLOs, error budgets, and burn rates separate SRE preparation from DevOps preparation.
- Higher levels are judged by the scope of the problems they solve, not by different questions.
- Confirm the exact rounds with the recruiter, then prepare by round.

GET YOUR FREE
Coding Questions Catalog

$98

$98

$197