System design interview question
Design a Log Aggregation System (ELK-style)
A log aggregation system ships logs from thousands of servers into a centralised store, parses and enriches them in flight, indexes them for fast search, and retains them across hot / warm / cold tiers for compliance and post-mortems. It must absorb bursts (incidents create traffic spikes precisely when the system is most needed), avoid logging itself to death, and answer dashboard queries in seconds across days of multi-TB indices. The design uses Filebeat/Fluentd agents, Kafka as a buffer, a parsing pipeline, time-based Elasticsearch indices, and lifecycle management to tier old data into cheaper storage.
Practice this question step by step with an AI interviewer that reviews your requirements, your architecture and your trade-offs, and scores the design.