Thursday 26 June 2014

What is Datamart ? why we need it? Types of Datamarts


Definition of datamart:–

Datamart is a subset of a Data Warehouse that supports the requirements of a particular department or business function.
Data Mart is specialized single-line of business warehouses. Data Mart collects data for a single department or a specific group of people.


Below are the reasons to create Data Marts :-

1. Provide users access to data the need to analyze most often
2. Provide data in a common format for a group of users
3. Improve end user response time
4. Provide appropriately structured data based on the requirements of end user access
    tools
5.Less complex implementation and setup
6.Reduced cost
7.Easier to support users


Types of Datamarts


There are three types of Datamarts mentioned below :-

1. Multidimensional Datamarts
2. Relational Datamarts
3. Networked Datamarts


Multidimensional Datamarts:-

Contain only numeric data
Maintain structure as data enters the framework
Only changed cells are updated and not the others
Contain sparsely populated matrices

Relational Data Marts :-

Contains numeric and textual both the data
Support star schemas/snow flake schema
Maintain structured data
Employ indexes while creating datamarts

Networked Data Marts:-

Data needs to be cleansed and standardized
Distributed data marts can be accessed from a single workstation
A standard metadata directory must exist in it
 

No comments:

Post a Comment