Data Warehousing - Data Marts

Data System Architecture

About

A data mart is a subject-or application-specific multidimensional schema build on the top of an Enterprise Data Warehouse.

Data marts are designed for a particular line of business and is an aggregation layer.

Data Mart

Data marts contain dimensional data (dimensions and facts). Facts can contain either atomic data and, if necessary, summarized data.

Data marts are designed and tuned for online analytical processing (OLAP) queries.

Type of data mart

  • Star Schema (Physical table on a relational Database)
  • Virtualized Star Schema (View on a relational Database)

When not virtualized, it must only store rolling history (ie limited history – not the entire history that is contained in the EDW).

Documentation / Reference





Discover More
3nf Sales
Data Warehouse - Data Models

In the data warehousing environment, you can find several relational data models. The staging area is a temporary layer where data can persisted in order to assist the data integration in the persistence...
Dw Layers
Data Warehouse - Layer (Architecture)

Name Properties Stage Layer Real-Time, CDC, continuous refresh DataWarehouse Layer / ODS Normalized, Data History, refresh: 2-6 daily Data Mart Performance, Access layer, Star schema, refresh: 1-4...
Data System Architecture
Data Warehouse - Persistent layer (Fundamental Layer | Operational Data Store - ODS)

A persistent layer is a data warehouse layer where data is persisted (ie never deleted). A persistent layer is also known as: a fundamental layer an operational data store (ODS) an enterprise...



Share this page:
Follow us:
Task Runner