Google·Software Engineer·Onsite - System Design / Architecture
- Design an end-to-end log processing system that ingests logs from many services and supports search and aggregation across the full pipeline.
- How would you handle log parsing and structuring, and what are the tradeoffs between schema-on-read versus schema-on-write?
- Walk through the query API layer: how do you support full-text search, time-range filters, and field-level aggregations, and how does a dashboard sit on top of that?
- How do you ensure reliability in the pipeline, specifically around at-least-once versus exactly-once delivery, and how do you handle deduplication?
- What are your cost and scale targets, and how does the architecture change as ingestion volume grows, say into the TB per day range?
“Big question, lots of moving parts.”