A graphical user interface enables you to easily create and document data flows and data flow templates in the Data Warehousing Workbench.
In the following cases, you can make full use of the advantages of graphic data flow modeling:
Application Scenario |
Advantages |
Top-down modeling of new data flows |
|
Structuring your implementation using data flows |
|
Organisation of existing models in data flows after upgrading to the current release |
|
Using data flows as templates to set modeling standards |
|
The data flow or data flow template is a standalone TLOGO object type (DMOD). Data flows and data flow templates can be transported and have a repository connection and document connection as well as a connection to version management.
Explanation of the Terms Data Flow and Data Flow Template
Data Flow
A data flow in the Data Warehouse displays a set of metadata objects in BW and their interrelationships. The relationships are displayed using transformations or are contained directly within the objects (as with MultiProviders, for example). The data flow specifies which objects are required for design time (modeling objects) and which processes and transformations are required at runtime (runtime objects), in order to be able to transfer semantically related data to BW, where the data can be consoldiated and integrated.
A data flow of object type DMOD can contain persistent and non-persistent objects. We define persistent objects as any objects that are already saved in the metadata tables in the database and which can also be displayed in the object trees of the Data Warehousing Workbench, for example, in the InfoProvider tree. A persistent object can be contained in multiple data flows and can therefore be used again in different data flows. We define non-persistent objects as objects that have been created in the data flow maintenance but only attributes such as object type and name have been specified for these objects. These objects have also not been saved in the database. Non-persistent objects can only be displayed and used in the data flow in which they were created. They also cannot be displayed in any other object tree of the Data Warehousing Workbench.
Data Flow Template
A data flow can be saved and used as a data flow template if the data flow only contains non-persistent objects. These objects are used as place holders to define the object and several other properties (technical name and description).
A data flow template describes a data flow scenario with all the required objects and can provide a scenario documentation, including information on why the objects are modeled in this way. Data flow templates are ideal for storing and documenting best practice modeling knowledge, which you can use to define the data flows. Data flow templates support the complex modeling of differentiated Data Warehouse layers in a Layered Scalable Architecture as well as fast modeling of simple standard data flows. You can also define your own templates to set standards for your organization.
SAP provides documented data flow templates that you can use in your BW system when implementing an LSA.
Integration into the Data Warehousing Workbench
Data flows (and data flow templates) are displayed in the Data Warehousing Workbench modeling screen with the symbol in a separate object tree. The data flows are structured like InfoProviders by using InfoAreas. The folders (organized by object type) underneath a data flow show you which persistent objects (including runtime objects and transformations) are contained in the data flow. Here the remaining objects for a persistent object are displayed downwards in the data flow (as in the other object trees in the Data Warehousing Workbench). These objects are displayed even if they are not contained in the data flow that is modeled here. This means that it is easier to include relevant objects in the modeled data flow.
In the object tree, the symbol in the object information column indicates whether a data flow has been created as a template. You can open the data flow maintenance screen for a data flow by choosing the context menu command Display or Change.
Note that the SAP data flow templates provided by SAP are only displayed in the object tree if they exist in an active version. The active versions are then displayed in the Business Information Warehouse InfoArea. For use in customer-specific data flows however, the SAP data flow templates do not need to be copied over from the content to the active version. You can display the SAP data flow templates in Data Flow Maintenance by pressing and integrate them into your data flow in the content version.
Data Flow Display in the Data Flow Maintenance
The data flow maintenance shows a data flow (or template) as a network with the following objects:
Modeling objects that contain other objects or are derived from them: for example, InfoSets, MultiProviders, semantically partitioned objects, aggregation levels, HybridProviders
Modeling objects that are used as the source and/or target for data transfer processes and transformations: for example, InfoCubes, DataStore objects, DataSources, open hub destinations
Objects that have source or destination objects: Transformations, data transfer processes
Other objects: InfoPackages that DataSources in BW have as a destination
Query and reporting objects do not belong to the data flow and are not displayed.
The components of the data flow are displayed as nodes (boxes) or connections (arrows). InfoPackages and modelling objects are displayed in the data flow maintenence screen as nodes. Data transfer processes, transformations as well as relationships between composite or derived modeling objects and the objects they contain are displayed as connection by default.
Nodes for persistent modeling objects (including transformations) are highlighted in blue, nodes for persistent runtime objects are highlighted in gray and non-persistent objects are highlighted in white. The object symbols for connections are displayed in white until an active version of the object becomes available.