OLTP (Online Transaction Processing) : Deals with operational data, where frequent DML(Insert/Update/Delete) operations happens, this is normalized system.
OLTP provides data to OLAP
OLAP (Online Analytical Processing) : Deals with historical & archival data, operations are
analysis and search of data used for reporting for purpose of data analysis and planning.
It is data archiving mining of OLTP data, OLAP is used for warehouses.
This required cost effective design for select statements not for insert/update/delete.
OLAP mainly required Facts, Dimensions & Measures for analysis & planning.
Fact - Product + Measures, Dimension - SalePerMonth, Measures - 30000$
Fact - Employee + Measures, Dimension - AgeWiseSalary, Measures - 2500$
So OLAP have demoralized design
There are two techniques of designing OLAP :
OLTP provides data to OLAP
OLAP (Online Analytical Processing) : Deals with historical & archival data, operations are
analysis and search of data used for reporting for purpose of data analysis and planning.
It is data archiving mining of OLTP data, OLAP is used for warehouses.
This required cost effective design for select statements not for insert/update/delete.
OLAP mainly required Facts, Dimensions & Measures for analysis & planning.
Fact - Product + Measures, Dimension - SalePerMonth, Measures - 30000$
Fact - Employee + Measures, Dimension - AgeWiseSalary, Measures - 2500$
There are two techniques of designing OLAP :
- Star Schema
- Snow Flake
Difference between them is that dimension could be normalized in Snow Flake.
