Chapter 7. Developing the BI Server Repository

Finally, we get to the heart of the OBIEE system, the metadata repository that resides within the RPD file. The RPD contains all information regarding the physical tables that are held in our database or data warehouse, whether this is from a single source or a heterogeneous set of sources. It stores their relationships, additional business logic, and the structure of how columns are presented via the frontend to the dashboard creators. As well as physical database tables, the RPD can also utilize other data sources, such as Essbase, Oracle OLAP, and Excel sheets.

The RPD also holds variable definitions, various security and cache settings, and drill/dimensional hierarchies that affect the end functionality of reports. All of these settings and metadata are used by the Oracle BI server choosing the content and structure of database queries when presented with a request via OBIEE Answers.

Once a system is installed and has gone live, a good amount of development time will be spent on the metadata repository. An RPD developer is given a large amount of freedom in modeling and enhancing physical objects in a way that can satisfy the most complex business requirements.

In this chapter, we will cover the development of a simple RPD, from importing tables in a database through to how those objects are presented to us when we move on to create an actual request. This will be carried out via the Oracle BI Administration Tool, which is the primary method of accessing and modifying an RPD file. This tool provides an inviting graphical interface for developing and administering the RPD. By the end of this chapter, you will be able to complete the major tasks associated with RPD development. We will also briefly describe some of the more advanced options that are available. This will give you a foundation on which you can investigate these tasks under your own supervision.

Prerequisites

The example we will be using stems from a Star schema design for a data warehouse. It is expected that you are already familiar with terms such as Star, Snowflake, Fact, and Dimension, and that you have an understanding of data warehouse schemas. The vast majority of development projects out there will use such a design as the basis for their BI projects. If you are not familiar with this, we would recommend studying one of the seminal texts on this subject beforehand - The Data Warehouse Toolkit by Ralph Kimball.

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
18.188.85.135