Additional Blogs by SAP
Showing results for 
Search instead for 
Did you mean: 
Former Member

SAP NetWeaver BW 7.30 provides you with a new metadata and transport object for BW data flows. This functionality enables you to persist the data flow views that you created with the ‘Display Data Flow’ function. Data flows are integrated as trees into the Data Warehousing Workbench. They are created and maintained in the same graphical design environment that you will be familiar with from transformation and process chain maintenance screens, for example.




A data flow provides a constant view of objects in the system. You can integrate existing objects and create new objects. The data flow maintenance screen can be used as a single point of entry for future development of your Data Warehousing objects and you can use the dataflow functionality to organize your projects.

Next I want to give you a short demonstration of some of the user-friendly modeling capabilities of data flow maintenance. I will also show you how  to model a data flow from a source to a target, based on existing objects or by creating new objects. I have modeled the following sales scenario: The USA data is sent from a DataSource via a DataStore object to the USA Sales InfoCube. The same setup applies to the Europe data, but the data comes from three DataSources. The USA InfoCube and Europe InfoCube are combined using a MultiProvider for reporting.  Let’s assume that some of the objects are already available and others have to be created from scratch.

Let's start in the data flow tree of the Data Warehousing Workbench. We create a new data flow using the InfoArea context menu and enter the name and description of the data flow. We save the empty data flow and the object appears in the tree. Then we drag and drop the icon of the object that we want to create from the toolbar to the data flow maintenance area. We start with the DataStore object, the InfoCube for Sales Europe and the MultiProvider.




We have now created three non-persistent objects. This means that the objects only exist locally in this data flow and not in the database. Now we draw two lines from the DataStore object to the InfoCube and one line from the InfoCube to the MultiProvider.




What happened? When you drew the line from the DataStore object to the InfoCube, the system realized that the line was a transformation and created a non-persistent transformation. When you drew the second line, the system created a data transfer process (DTP) because a transformation already existed. Further lines would result in further DTPs. When you drew a line from the InfoCube to the MultiProvider, the system created a simple line. This indicates a contained-in relationship, which is the only valid possibility in this case.

Next we used the undo and redo buttons that allow you to undo and redo your previous actions. The screenshots below show the result of the last four actions being undone. In this case, the three connections as well as dragging and dropping the MultiProvider are reversed.



To redo your actions, simply open the dropdown menu of the redo button and select the required action or press the redo button four times, for example.




Before we move on, we’ll assign a technical name and description to the InfoCube by choosing the change option in the context menu. Note that this option does not physically create the InfoCube in the system. The InfoCube is still a non-persistent object that is saved locally in the data flow.

The Sales Europe DataStore object exists already in the system. We add it to the data flow by choosing a command from the context menu of the non-persistent object:



Choose the DataStore object ‘Sales Europe’ in the value help. The system adds the object to the data flow maintenance screen and changes the background color of the node to blue. This indicates that the node represents an existing persistent BW object.

Next we create the Sales Europe InfoCube by double-clicking on the non-persistent object. The dialog box for creating InfoCubes opens. The DataStore object is selected as the InfoCube template because it is connected to the InfoCube. The description and technical name of the non-persistent object are applied to the InfoCube.




Activate the InfoCube and create the transformations and DTPs by double-clicking on the appropriate connections. The source and target are already specified. You simply need to maintain the details. Activate the objects.

Now we need to integrate Sales USA (already exists in the system) into our data flow. We start by dragging the existing Sales USA InfoCube from the InfoProvider tree and dropping it onto the MultiProvider. The existing InfoCube is added. Since the target is the MultiProvider, the system creates a connection between the InfoCube and the MultiProvider, based on the algorithm explained above.




To integrate the entire data flow of the Sales USA InfoCube, expand the existing data flow of the  InfoCube by choosing the menu entry ‘Use data flow of object’. A dialog box opens. Specify whether you want to use the data flow upwards, downwards or in both directions. The data flow is shown in a separate window and can be added to the data flow maintainance by pressing the confirm button.




The objects of the data flow are now integrated in our data flow. Double click on the MultiProvider to open the MultiProvider maintenance screen. Here the source objects are listed due to the connections and are already selected in the MultiProvider maintenance. Simply complete your MultiProvider and activate it.




Connect three European Sales DataSources to the DataStore object by selecting multiple DataSources in a Data Warehousing Workbench object tree and dropping them onto the DataStore object. The transformations are automatically created (as mentioned above).




The Sales data flow scenario can be completed by drawing links for the DTPs and maintaining the details for transformation and DTPs.

This document gives you a first impression of what is possible with the new data flow maintenance. Note that this is only part of the functionality. Some additional features are listed below. A very important function is described in a separate document: Creating and Using Data Flow Templates.

  • If the data flow is collected for the transport, all relevant objects are also collected.
  • You can copy data flows with the data flow copy wizard by performing a simple copy or a multiple deep copy.
  • You can use the create data flow wizard to generate parts of your data flow.
  • Semantically partitioned objects are supported in the data flow maintenance
  • You can collect and integrate all DTPs and/or transformations that belong to objects in the data flow by selecting a single option. For example, if you start your data flow with a list of InfoProviders and you want to display all transformations that connect objects to other objects of this object set.
  • …and many more features.  Try it out!

You are interested in more BW 7.30 news? Check out the SAP BW Developers SDN Blog Series Accompanying the BW 7.3 Ramp-Up Phase