SQLMentor // glossary

Data Warehouse

A data warehouse is a database optimized for analytical queries and reporting across historical data from multiple sources, as opposed to an operational (OLTP) database optimized for fast, individual transactions.

Data warehouses typically use denormalized schemas (like a star schema), columnar storage, and heavy indexing/partitioning tuned for large aggregate scans rather than single-row lookups. Data usually arrives via an ETL (extract-transform-load) or ELT pipeline from one or more source OLTP systems, rather than being written directly by end-user applications.

The distinction matters for schema design: normalize for OLTP correctness and write speed; denormalize for OLAP read speed.