Reconcile Data Discrepancies Using Delta Load Levels

  • by Brian Hether, Manager of Data Warehousing, The Washington Post
  • Opesh Gupta, Sr. BW Developer, The Washington Post
  • January 1, 2008
If the revenue from the total billed orders plus the total unbilled orders in your source system doesn’t equal the total amount of order revenue you have in BW, a balance control report can help you reconcile the two. Learn how to create and use one using delta load levels.
Key Concept

Using the delta load level for data reconciliation instead of using all the data shortens the amount of time and work required because the system only has to check the data that has been added, changed, or deleted since the last data update.

When reviewing reports, the business side may encounter a situation in which two reports based on the same calculation provide different results. For example, query A from SAP BW says that you have total revenues of $1,000,000 for 2007, but report B from R/3 says you have $1,005,000 for 2007. Which one is correct?

It can take significant effort to explain this to the business, especially if you don’t have systematic, automated routines in place. The two systems can have different key figure values for many reasons, including that BW didn’t load the data correctly. For example, this could happen if the DataStore object (DSO) settings are not in line with the data (overwrite vs. aggregate).

To prove that BW loaded data correctly, you must be able to reconcile the data via delta processes while SAP BW is moving it from the Persistent Staging Area (PSA) to the DSO to the InfoCube. In our example, the total billed orders plus total unbilled orders in R/3 should equal the total amount of order revenue on the SAP BW side.

We developed a balance control process to reconcile the data as it moves from the PSA to the DSO to the InfoCube via delta processes. You can view the results via a query you run against a MultiProvider that brings together data from the PSA, DSO, and InfoCube. This process applies to SAP BW 3.5.

Brian Hether

Brian Hether is the manager of data warehousing at The Washington Post. He has 17 years of data warehousing experience, most recently with SAP NetWeaver BW, helping take the system live in 2007. He holds a bachelor’s degree in mathematics, a master’s degree in information systems, and a professional development certification in business administration from Georgetown University.

See more by this author

Opesh Gupta

Opesh Gupta works in the data warehousing solutions group at The Washington Post. He has 13 years of experience in designing, managing, and implementing software solutions on a blend of technologies from mainframe, client-server, Web development, and data warehousing. He has bachelor’s degree in engineering and a master’s degree in information technology.

See more by this author


No comments have been submitted on this article. 

Please log in to post a comment.

To learn more about subscription access to premium content, click here.