Data Federation, Horizontal Federation and Vertical Federation in OBIEE

Data Federation is the process of and implementation of integrating metadata from multiple data sources into a single metadata database to allow for the integrated reporting of the data from those multiple sources. In OBIEE terms, Data Federation involves bringing in the metadata from multiple sources into the physical layer of the RPD (BI Repository) and integrating the metadata into a single business model and possibly a single subject area.  In OBIEE, the data sources can be relational (OLTP databases or star-schema data warehouses), multidimensional (OLAP), or files (such as Excel or flat files); and the data can be of varying levels of aggregation in these sources.  This is one of the most powerful features of OBIEE.

When an Analysis (report) is run from Answers that uses data from these sources, the BI Server creates the appropriate SQL (OLTP) or MDX (OLAP) statements to retrieve the data from the appropriate source and then integrates the data for display to the user.  The user does not need to know what the source of the data is, how many sources there are, or how the data is retrieved.

There are two common patterns of how data can be federated in OBIEE for reporting – these patterns/methods are termed Horizontal Federation and Vertical Federation. I will explain them here.

Horizontal Federation involves setting multiple data sources for a common single logical table object in the Business Model and Mapping (BMM) layer of the RPD, such that the granularity of the data from the various sources is at the same level, and some columns of the single logical table come from one source, and some from another source(s) – basically each source adding columns (not rows) to the logical dataset.  A typical scenario for this pattern is where there is related data in multiple sources for a particular subject, but no single source holds the entire body of data for that subject – and this helps to bring all the descriptive data for the subject together into one.

Vertical Federation involves setting multiple data sources for a common single logical table object in the BMM layer of the RPD, such that the data for each columns could be coming from multiple sources, but at varying granularity levels – basically each set adding rows (not columns) to the logical dataset.  A typical scenario for this pattern is where data at an aggregated granularity is sourced from an aggregated OLAP data source or an aggregated OLTP data source, while data at a detailed granularity is sourced from the transaction level (OLTP or detailed-level star-schema) data source.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: