BW/4HANA In-memory Data-warehousing
The new reality for organisations happens against a quickly changing data landscape. Organisations and their customers access and consume/produce more data, more quickly than ever before. Fully structured, transactional data from organisation’ systems is joined by a new set of social and other customer data as well as large amounts of sensor and other machine data from the Internet of Things (IoT). And the data is everywhere — in our systems, in the cloud, in data lakes and other storage places, and in devices at our fingertips. The ability to create data analytics solutions quickly and effectively can mean the difference between an organisation performing or outperforming.
SAP BW/4HANA is the next generation data warehouse solution that offers fast deployments and rapid delivery of complex business analytics scenarios. BW/4HANA brings an opportunity for agile enterprise data warehousing (EDW) modelling and fast delivery of business analytics’ requirements for organisations seeking solutions in reporting, planning, dash-boarding, Big Data, and predictive analytics.
BW/4HANA like S/4HANA is optimised for the SAP HANA platform, including inheriting the high performance, simplicity, and agility of HANA. BW/4HANA delivers real-time enterprise wide analytics that minimises the movement of data, to allow a single, logical view of all data, including new data types and data sources by providing a logical data warehouse. It offers the ability to design a true logical EDW platform that can help organisations to run their EDW in a real-time design, but also caters to all possible scenarios that a modern datawarehouse expects…such as out of the box big data connectivity, smart data tiering, cloud readiness, advanced analytics, planning & consolidation, advanced processing like predictive, geospatial, text analysis etc
Does SAP S/4HANA replace SAP BW? What about embedded BW?
S/4HANA has tremendous reporting capabilities. Some reports which historically would run on SAP BW may be considered for reporting directly on the transactional system. This also means having the latest data at any time and avoiding the complexity of moving data from ECC towards BW.
If you are implementing S/4HANA and do not have a BW system in place, S4 definitely is a good starting point for the first day. Yet, the advantage of SAP BW (vs. a transactional system) is not limited only to the performance of an OLAP system. Most SAP BW systems include complex business transformations, consolidation of SAP systems (S/4HANA, Successfactors, etc.), non-SAP systems (machine, sensor data, etc.), structured and unstructured data sources (Facebook, Twitter, etc.). The increasing diversity of data sources also requires more agility to accommodate those changes in the your landscape.
As an Enterprise Data Warehouse (EDW), BW/4HANA consolidates data across the entire enterprise using a standardised but fully extensible data model, to support decision making. By defining common semantics, all data values get harmonised and the result is a ‘single point of truth’.
As of Netweaver 7.0, SAP ECC has an embedded SAP BW which has the advantage of consuming on-the-fly data that resides in S/4HANA, without any data replication. Yet, as mentioned earlier the benefit of running a separate instance of BW4/HANA includes the consolidation of data from multiple systems (also non-SAP). Going for the embedded BW can also become a disruptive element to the most critical priority of an ECC system: ensuring a production system up and running with a clear authorised access. Keeping Operational and Strategic systems separate should definitely be part of a your IT Architecture.
So what does BW/4HANA offer?
Key features of BW/4HANA Architecture
Lean application server
Runs exclusively on SAP HANA
SAP BW/4HANA Modeling Tools as integrated development environment
No Java stack required
Choice of analytics solutions
On-premise or in the cloud deployment options
Simplified Data Structures
SAP BW/4HANA brings along a simplified set of data warehouse objects and removes a lot of functionalities which are only required to overcome the limitations of traditional RDBMS (e.g. extended star schema/cube for OLAP reporting is in the HANA-optimised version no longer relevant). This reduces the complexity for designing, implementing and modifying the data warehouse environment.
The modelling structures have been reduced from 10 to the following objects:
SAP BW/4HANA limits the number of object types for data modelling. The Simplified Object Model eliminated InfoCubes and other legacy objects. Some legacy objects, like InfoObjects, are now optional. SAP BW/4HANA restricts objects to Open ODS views, CompositeProviders, and Advanced DSOs (and optional InfoObjects) to simplify the application development process.
Composite Provider: this object combines data from other InfoProviders and makes the data available for reporting
Open Operational Data Store View: this object acts as an interface between the SAP HANA objects (e.g. SAP HANA tables, views) and the BW modelling part for further modelling. It makes it possible to consume external data sources in BW without staging, combine data sources with BW models and physically load data from external data sources. They allow integration without having to create BW InfoObjects (field-based modeling)
Advanced DSO: they are required to persistently store the data.
Simplified Data Flows: from Layered Scalable Architecture (LSA) to LSA++
Traditional data warehouses are built upon the Layered Scalable Architecture(LSA). It is based on several layers with data being physically transferred (e.g. Acquisition, Transformation, Data Marts Layer). The LSA++ architecture offers a much simpler and smaller landscape which relies more on virtual objects and avoids duplicating data between layers without compromising the performance.
In LSA++, the first layer is the Open Operational Data Store Layer. Raw data is physically stored in BW with exactly the same structure as in the source system. Data can be extracted in batch mode or using real-time replication. The next layer is the Core Data Warehouse Layer where data can be transformed, cleansed and consolidated. This is where data from multiple sources (SAP & non-SAP) gets merged into one harmonised definition. The next layer is the “Virtual Data Mart Layer”. With HANA, those structures now become virtual and can combine data from other layers. Those virtual InfoProviders can be queried upon and offer more agility as opposed to the historical cubes where data had to be physically loaded. Depending on the needs to get data consolidated, the Core Data Warehouse Layer is optional
To summarise, LSA++ concept offers a framework for a more flexible data warehouse which deals easily with changes in business, technology and requirements.
Simplified Data Integration
BW/4HANA works with multiple sources of data including Hadoop or Cloud, whether structured and unstructured. This flexibility allows BW/4HANA to consume data in a variety of data types and formats. It integrates data from social media, third-party databases, and social media. It does this through a simplified four-component approach: operational data provisioning (ODP), HANA source system, big data source system, and a file source system.
Operation Data Provisioning (ODP): ODP provides a unified technology for data provisioning and consumption in BW. It acts as the hub for all data flowing into BW from external sources. It is used with SAP Landscape Transformation (SLT), ERP Extractor (SAP Business Suite), HANA Views and SAP BW.
With those two connectivity types, data can be made available in using real-time replication (HANA Smart Data Integration(SDI) based replication or via ODP )
Native SQL Access
You can easily expose BW/4HANA models as native HANA views, and you can build your own HANA models on top of the BW/4HANA models. You have the option of working with HANA in BW mode or in native SQL mode, or you can switch back and forth between the two.
When you activate objects within the BW/4HANA environment (e.g. ADSOs, Composite Provider), HANA views are automatically generated within the SAP HANA database. This allows using SQL logic on top of the generated views combined with data from native HANA (e.g. data fetched directly from S/4HANA). From an end-user perspective, HANA views can also be consumed by BI front-end tools.
Modern Interface
To align to the changes in the BW/4HANA, simpler and more modern interfaces are available for business users, developers and administrators, therefore enabling faster solution development to improve business agility. The interface is seamless with SAP HANA on premise, and Web-based for cloud deployments. Modelling tools are integrated with SAP HANA Studio(Eclipse)
High Performance — query at the speed of HANA
Apart from offering an in-memory data warehousing, BW/4HANA is designed to push an increased amount of OLAP functionality and computation to the HANA database as well as for the planning and data management elements. But the higher performance for BW/4HANA is not just about pushing the workload to the HANA engine. It is also about the greater integration to leverage the HANA platform capabilities within BW, for example, integrating data from a HANA model into a BW model. In addition, BW/4HANA provides the capability to leverage various advanced analytical libraries and algorithms for data mining or predictive analytic scenarios.
BW/4HANA is built on top of HANA and supports all ‘high-performance’ capabilities. Aggregates are no longer required, fewer indexes are needed, data loads and execution of queries run faster. This performance improvement is ensured by a push-down of complex BW operations and calculations to the HANA database (vs. the application layer).
Advanced Analytics
SAP BW/4HANA offers a strong foundation for real-time analytics. It provides a good interactivity between historical and live data residing in a diverse IT landscape. HANA specific libraries, like R-Script or custom HANA procedure are delivered to enhance the data with simulations, predictive analysis, etc.
Data Tiering: multi-temperature
BW/4HANA simplifies data lifecycle with a multi-temperature strategy by allowing automatic distribution of data to hot and warm storage areas. Cold data that needs to be saved for legal or regulatory requirements can be moved to Hadoop. BW/4HANA stores warm data that is rarely needed in FlashArray which is controlled by HANA. Hot data is stored in the SAP HANA in-memory database for immediate access and processing. Multi-temperature storage options lower the total cost of ownership, optimise management of data, and improve query response time.
Summary
SAP BW/4HANA is next-generation data warehouse that is optimised for the HANA platform, and inherits the high performance, simplicity, and agility. It offers a modern EDW for complex modelling solutions (e.g., reporting, dash boarding, planning, and predictive) on large datasets that can be designed in days instead of months, and introducing changes to business logic is a much easier process. Agility and simplified architecture lead to creating more efficient models, smaller data footprints, and, consequently, lower total cost of ownership.
Cloud readiness allows new prototypes to be built quickly and put into operation in days, not months, with low upfront costs. Out-of-the-box deployment to SAP HANA Enterprise Cloud, Amazon Web Services (AWS), or another provider make the cloud-deployment option a great choice to start with SAP BW/4HANA.
Useful links
SAP BW/4HANA Help Portal https://help.sap.com/viewer/product/SAP_BW4HANA/2.0.7/en-US
Comments