data warehouse

a massive, high-end library where information is carefully curated, organized, and archived for deep research

While they both store data, they serve fundamentally different masters: databases are for running the business; data warehouses are for analyzing the business.

DBs

• are OLTP (Online Transactional Processing) optimized for read/write operations
• normalized data to minimize redundancy. updating a record happens once in a single place.
• often stores data from a single app or process.

Data Warehouses

• are OLAP (Online Analytical Processing) optimized for read-only complex queries.
• de-normalized data, often in a star or snowflake schema. It prioritizes retrieval speed over storage, so it can group related data together.
• stores data from multiple sources

You can't use a DB for everything because if you try to run a massive analytical report on a prod DB you may lock the tables and consume so much CPU that the app might crash. A warehouse can be a separate playground to crunch numbers for hours without affecting the rest of the system.