Working with ODP Context - Calculation Views based extraction

You can transfer data from an SAP HANA database to SAP BW/4HANA using calculation views.

SAP HANA XS Calculation Views (including the obsolete Analytic Views and Attribute Views) are used. SAP HANA XS Advanced HDI calculation views are not supported.

The goal is to push down data preparation to the SAP HANA database using calculation views, before the results are loaded into SAP BW/4HANA. Calculation views are powerful data processing objects and include functionality that is not available in SAP BW/4HANA. For example, complex ranking, data intersection logic, and the ability to include custom SQL code in the calculation view to support complex scenarios. Calculation views can handle currency translation or the generation of restricted key figures, or key figures calculated before aggregation.

There are several benefits to using the context ODP_HANA including:

• A complementary loading approach compared to using the HANA Source System, which loads from database tables or database views.

• Direct loading from SAP HANA Calculation Views via DTP into SAP BW/4HANA InfoProviders (and as a consequence into the BW-managed schema)

• Support of mass data acquisition.

Be aware that in the context of SAP HANA Calculation Views, the operational delta queue (ODQ) is managed in the target system; that is, the consuming SAP BW/4HANA-system. As mentioned earlier, ODP is an SAP NetWeaver-based framework. In this context, the source is native SAP HANA and does not have any relation to SAP NetWeaver. For this reason, the queue needs to be managed in SAP BW/4HANA and so you need to access the tool there.

In the SAP HANA database, Analytic Privileges are used to define which users are allowed to view which data records for analysis purposes. The SAP HANA database is accessed using the SAP BW/4HANA database user SAP<SID>.To access a view in SAP HANA, this user requires the corresponding analytic privileges and the authorizations for the corresponding schema. In the BW system, the access by the user performing the extraction is controlled using authorization object S_RS_ODP_H. The authorization object can control access to SAP HANA package level and to the view level. It is recommended that you use the authorization object.

After migrating SAP HANA calculation views from the former SAP HANA 1.0 repository to the SAP HANA 2.0 Deployment Infrastructure (HDI), these SAP HANA calculation views are no longer available within the ODP_HANA source system context. This is because ODP_HANA does not support HDI and there are no plans to enable this support.

If you would still like to use calculation views for data loading this is still possible by using the new HDI XSA calculation views. To do this you need to approach this as follows:

First define an SAP HANA source system that points to the SAP HANA database schema that is assigned to the HDI Container where your calculation view run-time object is created. The run-time object is a database object known as a column view. The SAP HANA database schema where the column resides is automatically generated the first time you build a calculation view. Then, for this source system, create a DataSource in SAP BW/4HANA based on the column view of your calculation view that is generated during the build (activation) of the SAP HANA calculation view.

The most crucial restriction is that only SAP HANA XS calculation views are supported. Calculation Views created in XSA HDI environment are not compatible with ODP_HANA loading interface. SAP have no plans to close this gap. Considering this, the ODP_HANA context should not be used for new implementations and should only be considered if you already have calculation views built in XS Classic.

There are some additional restrictions for SAP HANA Calculation Views in relation with the ODP_HANA context:

• The technical name of the SAP HANA models is normally <package name> / <name of the view in SAP HANA>. However, there is a limit of 25 characters, and when this is exceeded a new technical name 2H* will be generated randomly.

• Calculated columns are only taken into account during extraction if they calculate before aggregation.

• The transfer of hierarchies defined for attributes is not supported.