Databases · Senior · Question 7 of 75
Why do row-oriented databases struggle with analytics queries over billions of rows?
- 1They don't support
GROUP BY - 2They can't compute aggregates
- ✓Correct answer: They read whole rows even when a query needs a few columns; columnar storage reads only needed columns and compresses them well
Explanation
OLTP (many small transactions) and OLAP (big scans) have different needs. Analytics often moves to ClickHouse, BigQuery, Snowflake or DuckDB.