Performance increased and Costs reduced

With a 15% share of the beer market, the Radeberger Gruppe is currently Germany's market leader. The group is part of the Dr. August Oetker KG family business in Bielefeld and a strategic business area of the Oetker group. The Radeberger Gruppe considers itself the guardian of German beer culture: the group is committed to regionality and emotionally-charged brands to cultivate the diversity of German beer, and contributes with its innovative business concept to advancing the German beer and beverage market. The Radeberger Gruppe uses SAP NetWeaver 7.0 to manage its business data. In recent years, the underlying database has grown relentlessly to approx. 1.1 TB as a result of day-to-day business, the integration of subsidiaries and, last but not least, projects from the individual specialist departments. Approximately 20 GB were added each month. This also caused the costs of storing and backing up this data to constantly rise. At the same time, performance in the SAP BW system was decreasing. It was whilst doing market research that the BI team of the Radeberger Gruppe came across PBS Software, the ILM specialist and SAP partner in Bensheim.

Solution

During initial meetings PBS presented the nearline archiving solution CBW NLS IQ. This solution enables you to archive the data in a Sybase IQ database and not in a separate ADK file as is the case with the traditional SAP archiving approach. The Sybase IQ database uses column-based technology and stores the archived data in compressed form. Thus the data volume is reduced on average by about 70 percent while the data is still available for reporting despite archiving. The result of Radeberger Gruppe's tests was that the data in the Sybase IQ database was read considerably faster than the data from the conventional database of SAP BW or from an SAP archiving solution. A key user test phase started for the nearline storage solution. About 50 users were asked to test the solution with regard to performance and data quality. Quickly all problems were eliminated, and production operation was started. Up to now, the eleven largest and most important data cubes have been archived with nearly 230 million data records.  To date, the Radeberger Gruppe's BI team has been able to observe an improvement in performance by an average of 40 percent in SAP NetWeaver BW compared to the traditional database. No special training was needed for the around 800 users working with the SAP BW system since the way they handle the data has not changed.

Conclusion

Among others, it is the snapshot function that makes the data access so fast. When using it, a copy of the master data and hierarchies is generated that is then also available in the Sybase IQ database. Transaction and master data together with master data attributes or also hierarchies that were replicated on a master date can be restricted selectively and directly on the IQ database. Due to this preselection, the resulting quantity of the transaction data is clearly smaller and is available faster for reporting in the BW system. Data loads that come into the data cube every day can also be archived in the IQ database by means of the request-based snapshot functionality without creating a time slice lock in the InfoCube.
The omission of the aggregates is another advantage. With regard to the already archived data cubes, this is on average 80 percent compression for the data that has been archived so far, and thus saves disk space, not to mention the administration effort that is saved.

Overview

Country: Germany
Industry: Consumer goods
Staff: about 4,800

Company
Germany's market leader with a 15% share of the beer market and with 16 locations.

Initial Situation
SAP NetWeaver 7.0 in usage. The underlying database had a volume of 1.1 TB and grew by 20 GB each month. This had a negative effect on the system performance and increased the costs for data retention.

Solution
Usage of PBS Nearline Storage solution while using the column-oriented Sybase IQ database technology (PBS CBW NLS IQ).

Benefit
Data volume is compressed by about 70 percent, archived data is still available for reporting. Performance improvement of about 40 percent.

"We were the first customer that went live with the solution. Therefore, we were able to roll in our ideas that were also implemented already during the development process."
Sebastian Smolkiewicz, Senior Consultant, Radeberger Gruppe KG