Creating HybridProviders

Due to its 'hybrid' architecture, the HybridProvider provides faster access to data when you analyze data in queries. "Hybrid" means that historic data is combined with a small set of new data. This ensures a high level of system performance, for example when analyzing data in a query that contains real-time data.

There are two types of HybridProvider: HybridProviders based on a DataStore object and HybridProviders based on direct access.

The HybridProvider based on a DataStore object is a combination of DataStore object and InfoCube. The newest data here is in the DataStore object. The InfoCube serves as an aggregate. By storing the InfoCube data in the BW Accelerator, you indirectly index DataStore object data there too.

The HybridProvider on direct access is based on a combination of VirtualProvider and InfoCube. The newest data is not replicated into BW. It is read directly in the source system using the VitualProvider instead.

1. Follow the general procedure for creating an InfoProvider. All HybridProvider-specific settings are described below.

   More information: Creating InfoProviders

2. On the screen for creating the InfoProvider, you can choose whether the data of the relevant InfoCube is stored in the BW system or in the BW Accelerator.

   More information: InfoCube with Data Persistency in the BWA

3. Choose the HybridProvider type: Either based on a DataStore object or direct access, that is, based on a VirtualProvider. You can create a new object or choose an existing one. You should choose an existing object if you want to extend an existing InfoProvider into a HybridProvider with environment (possibly data, transformation rules, queries, and so on).

   Note

   If you choose an existing DataStore object (DataStore object or VirtualProvider), it is no longer an independent object. It can only be used as part of the HybridProvider.

4. Choose Create.

5. The remaining procedure is the same as that for creating DataStore objects and InfoCubes (VirtualProviders are created in the same ways as InfoCubes). See:

   • Creating DataStore Objects

   • Creating InfoCubes

   When you create new objects, the HybridProvider initially does not contain any InfoObjects. If the underlying InfoProvider already exists, the HybridProvider contains these InfoObjects and can be extended to include further ones. The associated, required InfoCube is created in any case.

   There are two views for modeling a HybridProvider based on a DataStore object: A DataStore view and an InfoCube view. You can switch between the two views by choosing  Settings  View . If you add or delete InfoObjects in one view, the system automatically applies the change to both objects.

   Both objects contain the same InfoObjects. There is one exception to this: The DataStore object can contain additional InfoObjects that are not included in the InfoCube. In the InfoCube view, you assign these InfoObjects to the Contained in DataStore Only folder. These InfoObjects are no longer available in the Query Designer.

   HybridProviders based on direct access do not have two views because the two InfoProviders in question are identical.

   Note

   Note that key figures with the aggregation MAX or MIN cannot be used in HybridProvider queries. For this reason, you receive a warning if you attempt to add one of these key figures to a HybridProvider.

6. Save the HybridProvider and activate it. When HybridProviders based on a DataStore object are activated, the data transfer process, transformation between the two objects, and the associated process chain are generated at the same time.

   Note

   If the system does not generate the relevant objects automatically, you can generate them manually. To do so, choose  Environment  Generate Data Flow Environment .