Home Page for the TeradataForum
 

Archives of the TeradataForum

Message Posted: Sat, 09 Nov 2002 @ 16:25:11 GMT


     
  <Prev Next>   <<First <Prev Next> Last>>  


Subj:   Re: 3NF vs Dimensional modelling
 
From:   Roth, David W

Kumaran,

Your definition of "Data Warehouse" is too vauge.

There is a set of tables that defines the core business data modelled in 3NF to accurrately reflect the business data and relationships, and make for easier and more performant ETL loading. We refer to this as the Data Warehouse. This data can be used to answer any question at any time (based on the data loaded, of course).

Data Marts created off of this data tend to use different modelling techniques, i.e. star schemes because they are being designed to answer very specifc business and reporting needs. While these marts can be physically located on the same Teradata box, and often are for performance and ease of ETL, they are generally not viewed as part of the "data warehouse" but rather data to support some application or applications. Note that sometimes these marts are just views (a logical reuse of the base data) but many times are physically separate structures.

This appears to be a confusing issue to many: the difference between the physical box(s), the Teradata rdbms running on the box, the Data Warehouse set of tables, and then any other databases and data that for specific reasons are manifested in the same Teradata insatnce as the Data Warehouse data. It is important to understand the distinctions though, because different modelling strategies are used for the different catagories.


Regards,

Dave

David W. Roth
Teradata Advanced Certified Professional
TERADATA, a Division of NCR



     
  <Prev Next>   <<First <Prev Next> Last>>  
 
 
 
 
 
 
 
 
  
  Top Home Privacy Feedback  
 
 
Copyright for the TeradataForum (TDATA-L), Manta BlueSky    
Copyright 2016 - All Rights Reserved    
Last Modified: 15 Jun 2023