What is a Data Mart?

Data Mart is a repository of data to meet particular demands of a specific group within an organization as a subset of a data warehouse.

A data mart is a pattern specific to a data warehouse environment. Data warehouse has an enterprise quality depth but data mart is for a single department. Each data mart is dedicated to a specific business function. This subset of data may span across many functional subject areas. It is common for multiple data marts to be used in order to serve the needs of each individual business unit. In other words, a data mart contains repository of summarized data for analysis within an organization, like, the sales department. Organizations usually build data warehouse, data mart to make data readily accessible, avoiding queries which are too complicated or resource-consuming.

While the transactional databases are designed to be updated, data warehouses or marts are read only. These are designed to access large groups of related records. Data marts specially designed to improve the end-user response time. So a data mart is a condensed, focused version of a data warehouse which reflects the process specifications of an unit within an organization. A dependent data mart is a subset of a larger data warehouse. The related term is spreadmart. We need a Data mart to :

1. Easily access the frequently needed data
2. Speed up analytical queries by reducing the volume of data to be scanned
3. To give a structure to the data
4. Partition data to impose access control strategies
5. Segment data into various hardware

The data mart, can be a small data warehouse that draws all its data from a larger master data warehouse through an ETL process. Data warehouse is a centralised system whereas data mart is decentralised. Data mart has short life than data warehouse and are project oriented.

This Article Has Been Shared 2175 Times!

Facebook Twitter Google+ Pinterest