On-callHardoc-g024

Subject Latency spikesLevel Senior–Staff~40 minCommon in Reliability & on-call interviewsIndustries Technology, Software development

Question

Your product catalog service (Go, Redis cache in front of Postgres) serves 40k rps at peak. At 09:02 a Redis cluster failover completes after a node crash; the new primary comes up with an empty keyspace. Immediately, p99 on GET /product/{id} spikes from 25ms to 9s, Postgres CPU pegs at 100%, the Go service's goroutine count balloons from ~2k to 200k, and you start seeing context-deadline-exceeded errors. Redis hit rate is near 0% and slowly recovering. No code deploy occurred. Triage and mitigate.

What a strong answer looks like

Stop the bleeding first (mitigate), then form hypotheses from real signals. Separate root cause from symptom, communicate status as you go, and close with what prevents a repeat.

Learn the concepts

Diagram & narrate the incident

Loading whiteboard…

Run or narrate your approach, then ask the coach.