For all of you NetWeaver Consultants out there that focus on SEM-BPS, BI-IP or any other SAP Business Planning Solutions and are thinking of moving into the world of BPC, there is a whole new set of terminology out there which may or may not be familiar to you.
The objective of this blog is to help all of those folks out there like myself that have an SAP NetWeaver Pedigree and want a starting point to begin learning and understanding SAP’s new Business Planning and Consolidations application (formerly OutlookSoft). Remember that Business Planning and Consolidations is based on the Microsoft Platform and the newest release BPC 5.1 is based on Microsoft SQL Server 2005.
Since SAP BW 1.2A we have been supporting something known as MDX, or MultiDimensional eXpressions. MDX is a querying language similar to SQL that is used to query MultiDimensional data objects or InfoCubes in our case. MDX a standard now that was developed by Microsoft has a different structure than what we are used to in NetWeaver BI, however this diagram below (from help.sap.com) was created awhile back to help us correlate Microsoft objects with SAP objects (see figure 1.1).
Figure 1.1 – Mapping Metadata with Microsoft
This diagram is a good starting point to understanding some of the same terms used and leveraged within SAP BPC and how they correlate to NetWeaver. I really liked this diagram and how it illustrates side by side the terminology of Microsoft and the terminology of SAP so I went ahead and built some new diagrams that attempt to do something similar.
Below are a series of diagrams which draw dotted lines between SAP BPC 5.1 objects and SAP BI objects. These correlations are my own and have helped me understand quickly SAP BPC and the environment in which it runs.
Figure 1.2 – Mapping the Tools and Services within the Platforms
From within SQL Server 2005’s toolset, SQL Server Management Studio can be installed as a workstation component. This tool can be used to view existing Databases, their corresponding tables, views stored procedures etc. From within the ABAP Data Dictionary you can similarly view all of the tables, views and structures available within your NetWeaver installation. Along with being able to view relational objects via the SQL Server Management Studio if you connect to an instance of Analysis Services (Microsoft’s OLAP Engine) you can browse your cubes and see their corresponding dimensions very much like you can from the Administrator or Data Warehousing Workbench (tx RSA1) within NetWeaver.
As of SQL Server 2005 a new tool has been incorporated into their suite called SQL Server Integration Services or SSIS. SSIS is the second generation of what was called DTS or Data Transformation Services. SSIS allows a user to build a data flow using logic to determine a pathway of data based on different levers which can be pulled during the transformation process. Essentially, SSIS is used to load data into SQL Server data objects; tables, cubes, and master data.
SSAS, or SQL Server Analysis Services is Microsoft’s OLAP Engine (MOLAP based). IIS or Internet Information Services is Microsoft’s Web Server. As of SAP BW 3.0 every single SAP BI instance comes with a web application server built in.
At a very high level this begins to draw some correlations for those that have always lived in the NetWeaver world.
Let’s drill in a little bit to the “guts” of what we know as an InfoCube.
Figure 1.3 – OLAP Cube Structure
The diagram above is very self explanatory and is not too different from what we saw in Figure 1.1. The main points that I want to highlight here is that:
- A Property is a display attribute but SAP BPC has a feature by which you can flag a property as inApp which is the same thing as a Navigational Attribute.
- A Hierarchy in SAP BPC is not the same thing as a BI Hierarchy. A BPC Hierarchy is technically an Attribute within the master data table of a Characteristic InfoObject.
- Time InfoObjects are treated as Characteristic InfoObjects inside of SAP BPC.
The last diagram I wanted to go through details the underlying tables which make up an InfoCube.
Figure 1.4 – Tables that make up Cubes
Dimension Tables within NetWeaver correlate pretty much 1:1 with Dimensions added to a cube in SAP BPC. Be careful here because a lot of people use the word dimension very liberally and at times this can be a little bit confusing when you are trying to determine whether they are referring to a Characteristic InfoObject or a Dimension Table.
Within BPC there are three partitions which make up what we would think of as fact tables.
- Fact Table – This table corresponds to the F-Table in SAP NetWeaver
- Write Back Table – This table correlates functionally to an open request in a SAP NetWeaver InfoCube. When data is written from a planning application in SAP BPC this information is written directly to a separate table. Within NetWeaver this write back is performed by separating data in the F-table by a request ID and writing data directly to the most recent request in the cube that is open.
- Fact 2 Table – This table acts as kind of an E-Table. When performance starts to drag in SAP BPC you can “Optimize” your Application (this is an Analysis Services Cube and a heap of contextual metadata) which will move your data from the Write Back Table (this is a ROLAP partition) into the Fact 2 Table (a MOLAP partition). This is done specifically to improve performance
Hopefully these “dotted lines” will help clarify some of the confusion with terminology for the savvy NetWeaver BI folks out there that are getting up to speed on SAP BPC. Shortly I will blog about the BPC Admin Console and explain in NetWeaver Speak what the different objects are and how they relate to one another.