Oracle Business Intelligence Enterprise Edition Obiee Data Source Access With Physical Layer

What does a business intelligence system made of?

Concepts of Business Intelligence is really simple:

  • Data sources – data used by business intelligence applications for a specific business purpose (e.g. sales data, marketing data , expense data which might reside in differnt system of the enterprise.)
  • Data Model – a meaningful organization of the data from the combination of the data sources related to a specific business subject.
  • Analysis applications – how the end user or analyst see and manipulate the data model, might contain graph, drill-downable data table, different format of reports.
    Business Intelligence layersOBIEE data Sources Layer

The data sources Layer – the OBIEE physical layer

Physical layer of OBIEE usually define where is the data and how the data could be accessed. The end user and data modeler do not have to has any knowledge of the real data sources , The physical layer act like a magic portal, through which data could be retrieved. physical layer does not contain any actual data , It serve as data agent for the backend data source from different database system.

Physical layer can also store the usage policy of the backend datasource through the connection pool object. for example you could define a concurrent connection of maximum of 5 for a busy data source which will confine the concurrent query of 5 user at most to prevent BI query from using too much of the backend processing power.

Physical Layer is abstraction of the existing data sources /datamart

Unlike the traditional approach of having to use ETL tools to move data from different data sources into the central data analysis model, OBIEE could consolidate the existing data sources into one virtual datamart for business model requirement.When you issue any BI Query using OBIEE. The BI Server decompose the query into sub-query that will be submitted to different backend system which is defined in physical layer. The BI server combine all the execution result from different backend system and combine them into meaningful result and return the result to the end user.