InfoSource 

Definition

A non-persistent structure consisting of InfoObjects for joining two transformations.

Use

You use InfoSources if you want to run two sequential transformations in the data flow without storing the data again.

If you do not have transformations that run sequentially, you can model the data flow without InfoSources.

In the following cases we recommend that you use an InfoSource:

●     If you want to execute two transformations before data is written to the InfoProvider, you need an InfoSource to join these two transformations.

●     If you want to update data from several sources (more than one DataSource for one or more source systems), it is useful to consolidate the data in an initial transformation at InfoObject level and in a second transformation, perform transformations that are independent of the data and are valid for all data. This allows you to react in a flexible way to changes that are not data source dependent.

●     If you want to update data to several targets (InfoProviders), it is useful to perform the transformations that apply to all data in an initial transformation and in a second transformation, perform object-specific transformations. This allows you to react in a flexible way to changes that are data source specific.

●     You need an InfoSource if you want to read master data or use quantity conversions or currency translation. These three functions are only available in the second transformation on the basis of the InfoSource.

The following graphic shows how InfoSources are integrated in the dataflow:

Structure

In contrast to the InfoSource 3.x, as of Release SAP NetWeaver BI 2004s, an InfoSource behaves like an InfoSource with flexible update. See InfoSource 3.x.

The data in an InfoSource is updated to an InfoProvider using a transformation.

You can identify InfoObjects in the InfoSource as keys. In the transformation, the data records are then aggregated using this key.