Data Warehouse

A data warehouse is a system that consolidates structured data from many operational sources into a single repository optimized for analysis. It is the foundation of business intelligence: a clean, integrated, query-friendly store separate from the systems that run daily operations.

How It Works

Data flows into a warehouse from OLTP databases, applications, and external feeds through ETL or ELT pipelines, where it is cleaned, conformed, and organized. Warehouses typically use:

• Dimensional modeling with fact tables (measurements) and dimension tables (context), arranged in star or snowflake schemas.
• Columnar storage and compression for fast analytical scans.
• Massively parallel processing to query huge volumes.

Modern cloud data warehouses — Snowflake, Google BigQuery, Amazon Redshift, Azure Synapse — separate storage from compute, allowing each to scale independently and enabling elastic, pay-per-query analytics.

Why It Matters

A data warehouse gives an organization a single, trusted source for analytics. By integrating data from across the business and structuring it for queries, it powers dashboards, reports, and decision-making without burdening operational systems.

Its main limitation is rigidity: data must be structured and modeled before use, which adds upfront effort and handles semi-structured or raw data poorly. Data lakes and the lakehouse architecture emerged to address those gaps.

Related Terms

A data warehouse complements the data lake, is unified with it in the lakehouse, serves OLAP workloads, and is fed by ELT pipelines.