Today's decision makers, from frontline operational employees to analysts and executives, want to have a complete and up-to-date view of the business. This means acquiring and integrating data from different sources, providing up-to-the-minute information, overcoming departmental barriers, and delivering large amounts of historical data to perform strategic analyses.
For example, when a purchaser is in negotiations with one of your long-time vendors, a complete record of spending with that partner will determine your future purchasing power. If your reports do not reflect a unified view of business partners across all areas and systems within the enterprise, you inevitably will be negotiating with bad information — obsolete products, missing or duplicate vendor information, or an incomplete view of the relationships among vendors. This lack of a unified view of products and vendors — in other words, a lack of consistent master data — could cost you money.
How can you get an integrated
view of products and services into your systems and back out to users when
they need it? How can IT meet these challenges? What tools are there to
integrate, transform, consolidate, and distribute different kinds of data — whether it's transactional or master data, historical or up-to-the-minute?
This article describes how SAP NetWeaver addresses these data integration and distribution challenges, specifically in the context of immediate and accurate master data delivery. SAP customers already using SAP Business Intelligence (SAP BI) have crucial reporting capabilities for historical data.1 By integrating your current SAP BW 3.5 system with other components of SAP NetWeaver — SAP Master Data Management (SAP MDM) and SAP Exchange Infrastructure (SAP XI) — you can provide information consumers with unified data at the right time to increase their insight into the enterprise.
Expanding Your Approach to Data Acquisition: ETL and EAI
In many companies, data is fragmented and spread across dozens, if not hundreds, of databases and applications. To be useful, data must be integrated, standardized, synchronized, and enriched, typically through ETL (extraction, transformation, and loading) processes. SAP BI provides a broad set of ETL capabilities that support data extraction. With SAP BI's open interfaces, data can be loaded from virtually any source, and can handle the huge amounts of transactional data typical of the enterprise landscape.
Nowadays, though, companies are looking beyond traditional ETL and expanding the concept of data acquisition to include enterprise application integration (EAI). EAI enables applications to talk to each other and deliver information immediately. SAP XI, as an integration broker, supports the concept of real-time data acquisition, in which data is updated in the data warehouse or the operational data store in real time.
Of course, you don't always need
all data in real-time. Now, with SAP NetWeaver, companies can provide
up-to-the-minute data for immediate decisions, and deliver large amounts of historical data for those performing strategic analyses — without redundant data for each type of user. This is true whether you're dealing with transactional data or master data.
For a unified and synchronized picture specifically for product and vendor data, SAP MDM works in concert with SAP XI to provide consistent distribution of master data objects into other systems — both within the company and across company boundaries. Thus, SAP NetWeaver provides a unique platform that offers complete information integration.
How SAP NetWeaver Brings Consolidated Master Data Into Your BI System
SAP XI is the integration platform between SAP MDM and other areas of your solution landscape, such as SAP Enterprise Buyer Professional (SAP EBP), SAP R/3, and non-SAP systems, through the exchange of XML messages. The integration of SAP XI and SAP BI allows you to use SAP XI to send data from various sources, including SAP MDM, to SAP BI. SAP XI distributes (routing), transforms (mapping), and transports the messages that are exchanged between the systems.
When you use SAP XI to support real-time data acquisition, the data warehouse and operational data store in SAP BI can act as one more subscriber to the real-time data distributed by the integration broker. When you connect your current SAP BW component to SAP XI, as shown in Figure 1, SAP BI becomes a recipient to the data routed by SAP XI, including the master data from SAP MDM. In other words, SAP BI becomes a master data client. Achieving this requires three basic steps.
| Adding Master Data Acquisition and Distribution to Your Current SAP BI Capabilities
3 Steps Toward Master Data Distribution to SAP BI
Suppose the master data administrator or manager has created a new business partner object (BUPA) for one of your vendors on the master data server as part of the Central Master Data Management scenario (see sidebar).
This object contains the vendor's name, addresses, and credit and bank information, and it needs to be distributed to its respective master data clients. We want to define SAP BI as one of those clients — that is, as a recipient of the master data record (see Figure 2). Which
steps does the data administrator have to perform to distribute and receive master data in SAP BI?
- Identifying related Business Content Objects in SAP BI and generating the corresponding XML interface in SAP BI that will receive the data from SAP XI. SAP XI relies on exchange of XML messages for integration.
- Designing the inbound message interface in SAP XI's Integration Builder (including profile mapping) that will map the MDM object
structure to the business partner
and vendor structure in SAP BI.
- Generating the proxy object in SAP BI. With this proxy, SAP XI will be able to deliver the right messages to SAP BI via an appropriate service exactly once and in the correct order.
How Does SAP MDM Work?
In local systems within your landscape, such as SAP R/3, different local vendor numbers in different applications often represent a single business partner. For example, one SAP MDM scenario, Central Master Data Management, provides an environment where you can gather and maintain master data on a master data server.
Consider a long-standing vendor that has seen you through mergers, acquisitions, and new subsidiaries. There will almost always be multiple identifiers across these various systems. SAP MDM can receive this data from any of its source systems, including R/3 and non-SAP systems. Key mapping information in SAP MDM associates all the various vendor numbers, for example (say, 5711 in one system and 6711 in another) as a single master identifier — say, 4711. When local vendor numbers are mapped to this master reference number, you have a consistent view of all your partner interactions across the business.
The relationship, in the form of key record mappings, for each business partner/vendor is stored in SAP MDM. SAP MDM then distributes the reference record to the master data client, while the original reference record data remain in SAP MDM.
As a result, you avoid duplicate data in master data creation, distribution, and, ultimately, in your reporting for more consistent views of products and vendors.
| Component View of the Master Data Integration Scenario with the
Main Process Steps
Step 1 — Create XML DataSources
in SAP BI
Once a master data object has been created on the SAP MDM server, SAP XI can send this information as an IDoc to the master data client. In our example, SAP BI has to receive the business partner number and attributes, the vendor number and attributes in various systems, and the key mapping information that will indicate the relationship between the business partner and various vendor numbers. Since each receiver expects the data in a format according to its data model, the data must be split into several messages. ERP clients of
the SAP MDM server get their data
via IDocs. However, a certain message structure must be built that reflects the particular data model of SAP BI, with its InfoObjects, which reflect the entities, and ODS Objects, which reflect business partner relationships (e.g., that a partner is also a vendor).
Any interaction with SAP XI relies on XML, which means that each SAP BI data model must have a corresponding XML interface. So the first step to make SAP BI the receiver of an XI XML message is to generate an XML-based function module from the data sources in BW. SAP NetWeaver provides the ability to easily generate and view these interfaces across the entire process, in preparation for master data distribution.
Step 2 — Create Inbound and Outbound Message Interfaces
with SAP XI
Based on the implemented RFC
function module (one part of the XML DataSources generated in Step 1), we now can proceed on the XI side. You've generated one structure in SAP BI, and the SAP XI Integration Builder can import the RFC interface description from it and build the necessary XI objects out of it.
The structure description of the interface must match the one generated in SAP BI. In Figure 3, the familiar BI elements are visible from the Integration Builder, along with all the mapping information needed. The message BW_Vendor shown here describes the data model; all the
fields that are part of the message interface — business partner texts
and attributes, as well as vendor texts and attributes — are now available here in SAP XI.
| Description of the BW_Vendor Message Interface in
the SAP XI Integration Builder
This message next must be sent
in asynchronous mode to an inbound message interface, so that the data
can be distributed to SAP BI.
Step 3 — Generate Proxy Objects
for Your Inbound Message Interface in SAP BI
To send data from MDM to SAP XI's Integration Server, which transfers this data to SAP BI at runtime, a third step is required to set up a proper proxy communication: generate a proxy object (see Figure 4) based on the message that will bring this XML information into SAP BI.
| The SAP BI View of the Proxy Object Based on the
SAP NetWeaver provides an entire Proxy Framework that offers many
services for communication between SAP NetWeaver components. To establish this same functionality in the context of master data distribution,
we have to generate a proxy object
on the SAP BI side that is capable of receiving the data structure of a certain XML message and turning it into an ABAP format.
Figure 4 shows the result of the generation process in SAP BI. Based
on the BW_Vendor message from
SAP XI (viewed here in the BI Proxy Administration screen), a method with an Import interface can be generated. Within this method the function modules of the XML DataSources must be encapsulated to allow the data flow
into the right direction — that is, in our case, to write it to the BI delta queue.
Once these three steps are complete, your system is basically ready for master data distribution. In SAP BI, the data is written to the delta queue. From there, the data administrator can collect the data using the usual staging methods for deltas in SAP BI and then post it to the data targets. A process chain then will hand over the data to the data warehouse in order to transfer it into InfoProviders, which make the master data available for reporting.
Additional Benefits of Connecting SAP BI and SAP XI
In addition to real-time possibilities and the ability to receive data from various sources, the integration of SAP BI and SAP XI also offers these advantages:
Greater options for transformation of message content between sender and recipient
With its EAI capabilities, SAP XI can now push data into the delta queue of SAP BI. Data transfer in BI generally takes place via a data request that is sent from SAP BI to the source system (pull from the scheduler). You can now also send the data to SAP BI using external control via a data push into SAP BI. Data transfer takes place via transfer mechanisms that are sufficient for Simple Object Access Protocol (SOAP) and are XML based.
What's more, for BI to accept data from MDM or other sources, the data must be transformed into a format that corresponds to the interface of the function module that is generated in SAP BI and used for data transfer. In this case, it's necessary to map the MDM data structure to the BI information model.
SAP XI's Integration Builder supports mapping, and helps you adapt values and structures of your message to the recipient with graphical models and a clear overview of the objects to be mapped (see the example below). In this way, you can transfer different data objects to an SAP BI system using interface mapping. This allows you to take data from another system, such as the MDM system, transform it into XML or another format, and bring it into SAP BI, where it can be transformed for reporting.
In the Integration Builder, the data structure of the message reflects the data model of SAP BI concerning business partners and related vendors. Several master data tables and ODS objects in BI are affected such as InfoObjects like 0BPARTNER, 0VENDOR, 0GN_VENDOR, and ODS's like 0GN_BP and 0CI_KMPA1.
Improved communication using proxies
Proxies are executable interfaces generated in the application systems for
communication with the SAP XI Integration Server. Proxies guarantee full
quality of service (exactly once and in order). SAP XI sends the message to
the BI proxy framework of the SAP BI system, from which the master data
will be written into the BI delta queue.
By making full use of SAP NetWeaver, including SAP XI and SAP MDM, companies that are currently using SAP BW can take advantage of the synergy between ETL and EAI in integrating and moving data. Drawing on the respective strength of both technologies, combined with master data management, companies can access, integrate, and deliver data from system to system using SAP NetWeaver's integrated information model as a unifying element.
And because not all end users require the same timeliness or "freshness" of data from your SAP BW system, with SAP NetWeaver, companies can provide up-to-the-minute data and deliver large amounts of historical data, depending
on whether a user needs to perform strategic analyses or real-time updates.
The synergies here help companies improve decision making through better use of data. With the three capabilities of SAP NetWeaver working together
in a single environment and using one metadata model, companies can easily combine historical information and real-time data to give users a richer context based on consistent data from a wide variety of sources.
For more information on SAP NetWeaver 2004, please visit http://service.sap.com/nw04. For detailed documentation on SAP MDM scenarios and distributing master data
to your SAP BI system, please visit http://help.sap.com/nw04.
1 This article assumes a working knowledge of SAP Business Information Warehouse (SAP BW), the
technical component of SAP BI. For more information on SAP BW, please visit www.sdn.com or http://service.sap.com/bw.
Claudia Weller is Product Manager for SAP NetWeaver at SAP AG. Before joining the
SAP NetWeaver team, she worked for five years in Product Management for SAP Business Intelligence. She has a background
as a business economist.
Rainer Uhle has been working for SAP
AG for over 15 years. He started as a
trainer for Basis Technology and Logistics
and soon became Curriculum Development
Manager for LIS/BW in the nKnowledge Management
Department. He is currently a member of
the Product Management Team for SAP BI.