On-callHardoc-g324

Subject Scaling limitsLevel Senior–Staff~28 minCommon in Concurrency · Distributed systems interviewsIndustries Technology

Question

Your event-processing service consumes a Kafka topic with 8 partitions via a consumer group. After a customer onboarding tripled event volume, consumer lag grows steadily and processing falls behind SLA. You scaled the consumer deployment from 8 to 24 pods to catch up, but lag kept growing and 16 of the 24 pods sit idle consuming nothing. CPU on the 8 active pods is high. There was no code change. How do you triage and what's the durable fix?

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.