Reviewed by Aditya Kumar · Last reviewed 2026-03-24
**Partition pruning**: Filter on partition column (e.g., dt) skips entire partition files—query reads only matching segments. **Scalability**: 1B rows partitioned by month = 12 segments; unpartitioned = full scan. **Columnar + partition**: Both reduce I/O; partition first...
Partition pruning: Filter on partition column (e.g., dt) skips entire partition files—query reads only matching segments. Scalability: 1B rows partitioned by month = 12 segments; unpartitioned = full scan. Columnar + partition: Both reduce I/O; partition first (coarse), then columnar (fine). Trade-off: Low-cardinality partition (e.g., boolean) = few files, little pruning. Best: Date, region—high filter rate.
Red Flag: Partitioning by low-cardinality (e.g., status with 3 values). Pro-Move: 'We partitioned fact_sales by (year, month)—94% of queries filter dt; scan reduced from 2TB to 40GB typical.'
According to DataEngPrep.tech, this is one of the most frequently asked SQL interview questions, reported at 1 company. DataEngPrep.tech maintains an editor-reviewed database of 1,863 data engineering interview questions across 7 categories.