SAP BW 7.3 and BWA 7.2 - Features and Overview

SAP BW 7.3 is currently in ramp-up. Select customers beta tested since Jul 2010. It is now generally available for customers..

Major benefits of SAP Netweaver BW 7.3 are:

  • Accelerated data loads - SAP has rewritten undelying code for data loads for optimized performance.
  • Performance improvement for BW Acclerator
  • Tighter integration with SAP Business Objects Data Services
  • Enhanced Intergration with SAP Business Objects Metadata Managements
  • Automated creation of Semantic Partitioned Objects
  • Graphical Data flow modeling and best practice data modeling patterns
  • Admin Cockpit integrated into SAP Solution Manager
  • Wizard based system configuration - Supports copying data flows, process chains etc.

Semantic Partitioning :

     We do partition the data in infocubes based on time or other characteristic to improve the query perforance. For example, say
we want to partition based on cal year, for this we need to create a infocube and transformation for each year and filter out the year in the DTP. So far all of these activities are manual and repetitive. 
    Semantic Partitioning allows wizard based partitioning of infocubes and DSOs. With SPO (Semantic Partitioning Object) the system automatically generates corresponding moedling objects like DTP (including respective filter criteria depending on the data target) and transformations as well as process chain.
   SPO object can even be embedded into a multiprovider. Partitions can be defince by customer coding (Business Add-Ins).


Graphical Data Flow Modeling:


Currently there it is not possible to visualize entire flow of a project during design phase. With this new feature, one can draw and visualize the flow before developing the objects. And it is integrated with the workbench so one can create actual objects using context menu. 
Customer can define template data flows which can be used for future projects. SAP also provides best practice templates.
Wizard based data flow modeling supports generation of related data flow objects including new infoobjects as needed.
One can copy existing dataflows leveraging a "Wizard Like" user interface. Another nice new feature is it now supports copying of existing process chains. 
Data Flow Migration Tool : It looks like BW 7.3 allows migration of entire data flow from 3.x to 7.x . Currently, one has to migrate each object (Datasource, Transfer Rules etc.) individually.

Accelerated data loads for DSO and Master Data :

DataStore objects
  • Support for database partitioning by time characteristics
  • Remodeling of DataStore objects
  • Performance enhancements:
    • Activation is changed from single lookups to package fetch of active table
    • Runtime option “Unique data records only” prevents look-ups during activation,
  • Performance gain for “package fetch” enhancement (lab test results)
    • Avg. 20 – 40% improvement, Max. improvement measured – 2.5x faster
    • Varies by data profile (#inserts/updates/deletes) and database platform
Master data
  • Optimized delta load for master data due to package fetch instead of single look-ups
  • Init-load performance runtime option - “insert only” prevents look-ups during activation
Near-Line Storage (NLS)
  • Integration into Data Mart Layer (Support of MultiProviders)
  • Selections on navigational attributes will be passed to NLS

Hybrid Provider (Real Time Business Intelligence):
This is an exciting feature for me. It certainly helps in displaying most current data in real time reporting scenarios. Hybrid Provider consists of a DSO, Infoobject and automatically generated data flow between these objects. The cool thing with hybrid provider is during run time it combines the mass historic data in the infocube with latest data in the DSO.
To explain this further, this means that the data in the new request that is loaded into DSO but not yet loaded into infocube is available for reporting. This is also applies for BWA so the data in the request that is not yet indexed in BWA is available for reporting.
With Hybrid Provider and BWA 7.2 it is possible to load data directly into BWA skipping infocube by switching off database persistency of infocube.

SAP BW 7.3 provides tight integration with Data Services which is a SAP's ETL tool.
SAP BW 7.3 supports Teradata database.

BW Accelerator 7.2:

 There are several new capabilities with BWA 7.2. To avail these one should upgrade BW to 7.3 verstion.
  • F4 Help - Right now it takes lot of time for F4 help of large master data characteristics. With BWA 7.2 F4 help come from BWA with significant performance improvement.
  • Hierarchy Support
  • Multiprovider Calculation Handling
  • Exception Aggregation (Min, Max, count distinct)
  • "BW Workspace" Analytic Indexes
Other Advanced Feature are,
  •   BWA based infocube - One can load data directly into BWA without loading it into infocube.
  •   BWA Indexes on DSOs - BWA 7.2 supports indexing Data Store Objects

Future Direction and In Memory Computing HANA 

   Future release of BW (some time in 2011) will support fetching data from SAP's ERP in-memory computing appliance HANA. Further in future BW supports HANAas underlying database, you don't need a separate database licence (DB2 or Oracle).
Also Read:
Click here to watch and download SAP Teched presentation on SAP BW 7.3 and 7.2 and

Labels: