Interview questions in Monitoring & SRE, page 5
Design automated canary analysis scoring
Track the golden-signal SLIs, compare canary to baseline statistically, weight and combine into a score with promote/rollback thresholds.
Client-side vs server-side feature flags
Client-side is fast and offline-capable but exposes flag logic and risks stale or leaked values; server-side keeps logic secret and consistent but adds latency.
Auto-rollback on failed blue-green cutover
Shift traffic gradually behind a smart router, use deep health checks plus real SLI monitoring, and auto-revert to blue on breach while blue stays warm.
Canary a shared downstream microservice
Route a slice of traffic to the canary via mesh rules, propagate context, and use distributed tracing to measure impact on upstream callers across the full path.
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.
Load vs stress vs soak testing
Load tests expected traffic, stress pushes past limits to find the breaking point, soak runs sustained load for hours to expose leaks.
Diagnose database CPU saturation under load
Find the expensive queries via the database's stats, check for missing indexes and full scans, then fix with indexing, query rewrites, caching, or read replicas.
Capacity planning for annual traffic spikes
Forecast peak from history, load-test to find per-unit capacity, use autoscaling with pre-warming and headroom, and protect with caching and graceful degradation.
Little's Law for capacity planning
L equals lambda times W, concurrency equals arrival rate times time in system; rearrange to size threads or concurrency for a target throughput and latency.
Front-end performance budgets
A performance budget is an enforced limit on metrics like Core Web Vitals and bundle size, checked in CI to fail builds that regress.
Why tail latency (p99) matters
Averages hide the worst experiences, and fan-out amplifies tails so most requests touch a slow path; causes include GC pauses, queueing, contention, and noisy neighbors.
When horizontal scaling is the wrong fix
Adding instances fails when the bottleneck is a shared resource like one database, a lock, or a queue, so more instances just add contention; investigate where time is actually spent.
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.
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.
Reliability patterns for queue-based job processing?
Retries with backoff and jitter for transient faults, dead-letter queues plus a poison-message limit, idempotent handlers and visibility timeouts.
Mitigating risk from an unproven external dependency?
Timeouts and circuit breakers to fail fast, bulkheads to isolate resources, fallbacks or cached/degraded responses.
Zero-downtime index migration on a hot table?
Build the index concurrently to avoid table locks, run off-peak with monitoring, and keep it reversible since dropping an index is cheap.
Replication and consistency for active-active regions?
Choose per data class between synchronous (low RPO, higher latency) and async replication, address write conflicts, and reason via CAP and PACELC.
Design a chaos experiment for a payment dependency?
Define a measurable steady state, hypothesize it holds when payments fail, limit blast radius to a small traffic slice, and auto-abort on SLO breach.
Reliability paved roads for an internal PaaS?
Built-in observability, safe deploys with health checks and rollback, sane timeouts/retries/limits, and SLO tooling.
We are hiring for this. Every open role lists the topics its interview covers, so you can prepare for the real thing rather than guessing.
See open roles