Data warehousing

A data warehouse is a central store of an organisation's historical data, built for analysis rather than day-to-day transactions.

• Operational databases handle daily transactions; the warehouse consolidates them for decision support.
• ETL: Extract data from source systems, Transform it (clean, standardise, integrate), Load it into the warehouse.
• Warehouses are often denormalised for fast querying of huge histories.
• Data marts are department-focused slices of a warehouse.
• Feeds dashboards, trend reports and data mining.

• ETL — extract, transform, load.
• Denormalisation — trading redundancy for query speed in analytical stores.