SRE
205 bites tagged SRE — interview questions with model answers, and 60-second explainers.
Federating reliability ownership to product teams
Build a self-service reliability platform (golden paths, paved roads), train teams and embed SLO/on-call practices, and govern with standards plus error budget… Whether you can scale reliability by enabling teams, not gatekeeping.
Capacity planning for a distributed cache
Track hit ratio, memory and eviction rate, throughput/latency, and connections; correlate with growth to forecast when to add capacity before the hit ratio or evictions… Whether you can pick the right cache metrics and forecast scaling.
Capacity planning for distributed stateful systems
Account for replication and cross-region network, IOPS and disk throughput, connection limits, partition/shard balance, and N+1 region failover headroom; validate with load and chaos… Whether you plan capacity beyond raw resources.
Managing a risky release with feature flags
Deploy code dark behind an off flag, enable for internal then small percentage, monitor metrics, ramp gradually, then remove the flag. Whether you can decouple deploy from release and roll out safely.
Automating actions on error budget burn
Use multi-window burn-rate alerts; on fast burn, auto-trigger actions like halting deploys, rolling back, or scaling, with escalating tiers. Whether you can operationalize error budgets with automation.
Safeguards for automated remediation runbooks
Add rate limits and circuit breakers on restarts, escalate to humans after N attempts, log all actions, and check for cascading failure before acting. Whether you can build safe auto-remediation.
Breaking tunnel vision during an incident
Call out the assumption, ask for disconfirming evidence, list parallel hypotheses, split responders to investigate them, and anchor on what changed and the data. Whether you can counter confirmation bias under pressure.
Key roles in incident response
Incident Commander coordinates and decides but does not fix; Communications Lead handles stakeholders; Operations/SMEs do hands-on diagnosis and remediation. Whether you understand structured incident command.
First-SRE 90-day plan at a startup
Listen and measure first, pick one high-impact service, define SLIs/SLOs and basic alerting, then show reduced toil or incidents to leadership. Whether you can introduce SRE incrementally and show value.
Designing shallow vs deep health checks
Shallow checks confirm the process is alive; deep checks verify dependencies; use shallow for liveness/load-balancer routing and deep sparingly to avoid… Whether you understand health-check semantics and failure amplification.
Rolling update vs blue-green deployment
Rolling replaces instances gradually with minimal extra capacity but mixes versions; blue-green runs two full environments for instant switch and rollback at double the cost. Knowledge of deployment strategies and their costs.
Writing high-quality postmortem action items
Good action items are specific, assigned to an owner, prioritized, tracked to completion, and ideally prevent recurrence rather than just detect faster. Whether postmortems lead to real change.
Calculating downtime for a 99.9% SLO
0.1% of 30 days is roughly 43 minutes of allowed downtime; healthy budget enables faster shipping while depletion slows or freezes deploys. Whether you can compute and apply an error budget.
Designing an error budget policy
Define SLO and budget, tiered consequences as burn worsens, a feature freeze on exhaustion, and concrete earn-back criteria. Whether you can make SLOs enforceable, not decorative.
Keeping a postmortem blameless after an admission
Acknowledge the courage, redirect from who to why the system allowed it, ask what guardrails were missing. Whether you can protect blameless culture under pressure.
Conducting a Production Readiness Review
Assess monitoring and alerting, capacity and load testing, failure modes and dependencies, on-call and runbooks, and rollback or release safety. Whether you can systematically gate a launch on reliability.
Embedded vs consulting SRE engagement models
Embedded SREs sit inside one team for deep impact but limited reach; consulting SREs advise many teams broadly but shallowly. Understanding SRE team topologies and scaling. Treating either as universally best.
A team keeps blowing its error budget. First steps?
Analyze where the budget is burning via SLIs and postmortems, validate the SLO and SLIs are sound, then partner blamelessly on the top fixes. Data-driven, collaborative incident reduction.
What is an error budget and how is it used?
The budget is the allowed unreliability (100 percent minus the SLO); track its burn, ship freely when budget remains, and freeze risky changes to focus on reliability when exhausted. Understanding error budgets as a release governor.
What is an SLO and how do you define API availability?
SLO is a target on an SLI, availability SLI is good requests over valid requests, you need labeled request counts over a window. Understanding SLI versus SLO and how to measure reliability.
What telemetry must a new microservice ship with?
Metrics for trends like p99 latency and error rate, structured logs for per-request detail, traces for cross-service causality. Practical observability literacy. Listing only host CPU dashboards and no request-level signals.
Golden signals for capacity planning
Monitor the four golden signals, latency, traffic, errors, and saturation, from day one, watching percentiles and saturation to forecast scaling. Core observability fundamentals.
Measure ROI of toil reduction efforts
Track toil hours, percent of time on toil, incidents auto-resolved, and engineer cost saved, then frame as ROI and risk reduction. Quantifying automation value. vanity metrics or claiming success with no baseline measurement.
Systematically reduce noisy alert toil
Inventory alerts, measure frequency, actionability, and time cost, then prioritize by volume times effort. Data-driven toil reduction. muting alerts blindly or automating before measuring which ones actually hurt.
Get SRE bites daily.
Five a day, five minutes, offline. With quizzes so it sticks.
Open testing — you’ll join as an early tester.