Problem 10 - Olympiad

A data warehouse uses bitmap join indexes: BJI1 on (d.year, f.product) and BJI2 on (d.quarter, f.store). Query GROUP BY d.year, d.quarter needs to combine both indexes. Which bitmap operation is required and what is the resulting compression ratio compared to materializing the join?

Correct: N/A

The GROUP BY needs rows that satisfy both dimension predicates, so bitmap AND is applied. Each bitmap index compresses to ≈1 bit per row, while the join materializes foreign-key integers (4 bytes), giving a compression ratio of 32:1 per index. Combining two such indexes yields roughly 8:1 overall saving.