Transcript Data Modelling in SeETL

SeETL
Demonstration 04
Data Modelling in SeETL
www.InstantBI.com
1/3/2012
Introduction
 History of the BI4ALL Data Models
 How to build data models in SeETL
 The documentation database
 Summary
History of BI4ALL Data Models
 In the mid 90s numerous people noticed similarities in data models
 Standard template models were tried by a few companies
 Sean Kelly brought out Vertical Package Solutions (VPS)
 VPS was sold to Sybase and that became Industry Warehouse Studio
(IWS). Sold on Sybase IQ and Oracle
 9/11 and the follow on recession caused Sybase to defocus on IWS.
There was little revenue in data models
 In 2006 Sean Kelly and Peter Nolan agree to create the “next
generation” of data models based on experiences
 2007 first sale closed to Key Work Consulting GmBH
 Version 1.0 is “views in a text document”
 Version 2.0 is “data model in an excel workbook”
 Version 3.0 was a vast expansion of coverage
 Version 4.0 is work in progress
 There is still not much money in data models
How to Build Data Models in SeETL
 This is a demonstration.
The Documentation Database
 A “side by side” database. Can be a different RDBMS.
 SQL Server is the only supported documentation database
 The views are created as tables
 Primary keys and RI constraints are added
 Most databases will not allow keys or constraints to be
created on views. In a DW you do not want RI constraints
turned on anyway. So it is very hard to document directly
from the production database.
 This can be re-engineered into a data modelling tool
 The data model is also loaded into the dictionary
Summary
 History of the BI4ALL Data Models
 How to build data models in SeETL
 The documentation database
 Summary