12.2    Openness

In the context of a data warehousing application, openness refers to flexibility with respect to inbound and outbound flows of data. From an inbound perspective, openness is directly associated with data provisioning methods and the span of connectivity with different sources. From an outbound perspective, openness refers to how easily data can be made available for consumption by different frontend tools.

Why is openness crucial these days? Let’s start answering this question from an inbound perspective. Data sources are no longer restricted to a specified format or originate from systems on-premise. Relevant data can reside in the cloud or in data lakes, be of structured or unstructured natures, and be provisioned via a streamed, replicated, or extracted method. This variety of data types, formats, and flavors requires several different data access methods that only a truly open application can cater for. This is the case for SAP BW/4HANA.

SAP BW/4HANA introduces four source system types to enable a flexible, open integration with data sources:

12.2.1    Operational Data Provisioning Source System Type

The ODP source system type is the data provisioning method used by SAP BW/4HANA for the SAP backend system. This means that ODP-enabled SAP DataSources can be consumed easily by SAP BW/4HANA. SAP is increasing the number of ODP-enabled DataSources as a replacement for the classic delta queue-based DataSources. The current list of ODP-enabled SAP DataSources can be found in SAP Note 2232584 (Release of SAP Extractors for Operational Data Provisioning).

12.2.2    SAP HANA Source System Type

The SAP HANA source system type is the data provisioning method that leverages SAP HANA enterprise information management (EIM)—and specifically the SAP HANA smart data integration (SDI) component. SDI includes several different source system adapters, resulting in a flexible and powerful data integration mechanism.

SAP HANA, Hive, third-party RDBMSs, SOAP, Facebook, and Twitter are just a few of the adapter types available via SDI. The SAP HANA source system type also enables optimized loading data processes and caters for scenarios from typical data extraction to direct access for data virtualization and real-time replication.

12.2.3    Big Data Source System Type

The big data source system type is a new option offered with SAP BW/4HANA that enables integration with Hadoop data lakes via the Spark SQL adapter. By making use of Open ODS views and the Spark SQL adapter, it’s possible to access and stage data lake contents with SAP BW/4HANA. Further innovation is already planned for this source system type. This includes SAP HANA analysis process, based on Hadoop, as well as automation of data flows between data lakes and SAP BW/4HANA.

12.2.4    File Source System Type

The file source system type enables the traditional data provisioning method of uploading content via file-based DataSources.

The four source system types are collectively referred to by SAP as simplified data integration offered by SAP BW/4HANA. In comparison with previous versions of SAP BW applications, SAP BW/4HANA offers more data integration options via fewer source system types (hence simplified). Figure 12.3 illustrates this simplification.

Simplified Data Integration with SAP BW/4HANA

Figure 12.3    Simplified Data Integration with SAP BW/4HANA

Now that we’ve discussed openness from a data inbound perspective, let’s shift to a data outbound perspective. The level of outbound openness of an application is directly related to how easily it can provide data for consumption by different frontend tools. This is especially relevant now that options for analytical and visualization tools are numerous and being created at an impressive pace, including open-source options. An application restricted to proprietary frontend tools can be considered an obstacle more than an enabler.

SAP BW/4HANA allows for automatic generation of SAP HANA views based on queries or any of its modeling objects (CompositeProviders, Advanced DSOs, or Open ODS views). This means that the data modeled in SAP BW/4HANA, including the calculation logic defined in queries, can be exposed to the SAP HANA database and consumed via SQL-based access. By doing so, SAP BW/4HANA enables flexible combinations of SAP BW models and SAP HANA native data sets. Further modeling activities can be performed with the deployment of logic applied via subsequent SAP HANA native models, known as SAP HANA views. The final models can be accessed easily via SQL-access-enabled frontend tools.

Looking at the road map for SAP BW/4HANA, openness will continue to be a driving principle. We expect to see tighter integration with Hadoop, an increased number of SDI sources, and enablement of further ODP DataSources for SAP backend systems.