When is denormalization typically used, and what is a common trade-off?

• A. Denormalization is used to enforce data integrity across tables.
• B. Denormalization reduces data redundancy.
• C. Denormalization is only used in data warehousing.
• D. Denormalization is used for performance optimization in OLTP/OLAP by reducing joins; the trade-off is data redundancy and potential update anomalies.

Answer: (D)

Denormalization is used to improve read performance by duplicating data so that queries can be answered with fewer or no joins. This is especially helpful in OLTP and OLAP workloads where fast data access matters, as denormalized structures allow a single table to cover what would otherwise require multiple joins. The trade-off is data redundancy and the risk of update anomalies, since the same information exists in multiple places and must be kept consistent. The other statements misstate what denormalization does or its scope: it’s not about enforcing data integrity across tables (that’s normalization), it typically increases redundancy rather than reduces it, and it isn’t limited to data warehousing.