Since 2012, it has been possible to install the SAP Business Warehouse (SAP BW) data warehouse using the solution SAP HANA (The High Performance Analytical Appliance by SAP) instead of a classic database. The expansion of the applicability of the SAP HANA product to allow its use together with SAP BW is positive news. This combination of the two SAP products allows data management through data warehousing mechanisms, while making available the incredible performance and flexibility associated with the latest data storage technologies.
Wanting to test this solution in practice, BCC (currently All for One Poland), together with one of the largest meat processing companies – Sokolow, and in cooperation with the leader in the server market – Hewlett Packard Poland, decided to perform a Proof-of-Concept (PoC) project to test the usefulness of this platform. This article summarizes the experience of this project.
Sokolow S.A. has so far been using SAP BW data warehouse version 7.02, installed on a powerful IBM iSeries platform.
For the purpose of the test, an additional data warehouse application server SAP BW version 7.3 based on the SAP HANA database platform provided by HP Poland was prepared. The contents of Sokołów SA’s data warehouse were migrated to this server. It became possible to report from the legacy system and from the new system and to compare the results of their operation.
| Element | IBM iSeries legacy server | SAP BW based on SAP HANA |
| SAP BW version | SAP Netweaver BW 7.02 | SAP Netweaver BW 7.3 on HANA |
| Database | DB2 UDB for iSeries V6R1 | SAP HDB 1.0 In-Memory Database |
| OS | i5/OS V6R1 (OS/400) | SUSE Linux Enterprise Server 11.2 |
| Processors | 5x POWER6 4.2 GHz | 8x Westmere EX Intel E7-4870 2.4 GHz |
| RAM | 80 GB | 512 GB (maximum use of 93 GB) |
| Drives | 36 x 141 GB | LOG: 2x Fusion-io ioDrive Duo 320 GB, RAID 0 configuration, DATA: 24x 146 GB disks 15k rpm, RAID 5 configuration |
Testing methodology
Analysis performance tests were conducted by analyzing the speed as seen “from the user’s side" and assuming the most flexible data warehouse architecture based on the HANA product. This ruled out the use of so-called “aggregates" to speed up reporting, which are commonly used in the SAP BW system used to date. The use of “aggregates" enables the execution of reports (without them, the time limit for a report was exceeded). The creation of new reports was burdened in the current system by the additional expense of adjusting, or creating new structures to speed up the execution of analyses.
One of the tasks posed to the HANA platform was to get rid of this additional administrative work hindering the free creation of reports by end users. Thus, the performance of the SAP HANA platform was analyzed without the additional optimization of analysis by means of data “aggregates." The measurements also did not separate the processing times on the database side from the times spent by the server to display the results found. Although comparing only data retrieval times would have presented better results for the new platform, it was recognized that what was important for the end user was the aggregate time from running the report to receiving the results. Such “business" time was therefore decided to be measured.
For data loading performance, the focus was on several of the largest data sets. Historical data loading times on the platform used to date were compared to those recorded in a system based on the SAP HANA product. The measured times were related to the number of data records loaded, which made it possible to compare loading acceleration for a single record.
| Report |
Execution time Existing server with data aggregation |
Execution time SAP HANA platform, without data aggregation |
Acceleration (multiplicity) |
| Profitability report for the period | 00:03:17 | 00:00:18 | 11 |
| Collection 2 | 00:00:51 | 00:00:03 | 17 |
| Collection 1 | 00:00:18 | 00:00:02 | 9 |
| Profitability report for the day | 00:03:42 | 00:00:07 | 32 |
| Comparison of valuations and base prices | 00:02:25 | 00:00:14 | 10 |
| Average acceleration (without optimization on the SAP HANA side) | 11 | ||
Measured results
Measurement results for reporting are in line with expectations. The data warehouse based on the new SAP HANA data access platform, despite the abandonment of data aggregation, turned out to be more than ten times faster, and, more importantly, the response times even for queries combining data from several data providers during the execution of reports were stable and always remained below one minute. It can be assumed that if optimization of data in “aggregates" had been applied, performance could have been increased by another order of magnitude.
| Area | Execution time/ Number of records Server legacy |
Execution time/ Number of records SAP HANA platform |
Acceleration (tuple) |
| CO-PA profitability analysis module data |
08:35:54 / 2,682,216 11,544 ms/record |
04:12:41 / 7,226,592 2,098 ms/record |
5 |
| Name | Description of business benefits |
| Accelerate SAP BW system performance in reporting | Acceleration on the order of 10 times, assuming no optimization of SAP BW based on SAP HANA (the lack of need for optimization provides other benefits related to improved self-service and less administrative effort). If one were to optimize the SAP HANA-based wholesaler, accelerations of up to 100 times could be achieved. |
| Increasing employee self-service for report creation | Enabling reports directly at the line-item level and not having to create performance-optimizing aggregates allows business users to create ad-hoc reports without having to interact with IT. |
| Revised approach to reporting | To analyze data, you no longer need to create rigid reports. You can use data analysis (e.g., the SAP BusinessObjects Explorer tool) to help you discover data in an area without creating reports. |
| Reduced costs of administering the data warehouse | Since data structures usually do not need to be optimized for performance when using the HANA platform, the data model is simplified. The data warehouse is simpler to manage, which reduces administrative costs. |
| Better collaboration of SAP BW data warehouse with SAP BusinessObjects reporting tools | Due to the increase in performance of the SAP BW data warehouse, you do not need to optimize queries for reports created in SAP BusinessObjects tools. This allows you to create generic queries on the SAP BW side at the data area level and create detailed analyses with SAP BusinessObjects tools only (without having to create queries in both systems) |
| Faster and more frequent analytics updates | Due to a significant increase in the efficiency of loading data into SAP BW (a measured fivefold acceleration), it is possible to make data synchronized with the source systems available earlier (especially important during month-end closing periods). In addition, it is also possible to increase the frequency of data loading into the warehouse in order to make data available as up-to-date as possible. In cases where data needs to be reloaded (e.g., data reorganization), system unavailability times can be minimized. |
Applications
Using the SAP HANA platform as a data storage layer for the SAP BW data warehouse has many advantages. In addition to the obvious increase in productivity (more than tenfold), it is important to simplify the administration of the data warehouse, making it possible to significantly increase employee self-service. Business users no longer need to work with IT support to create a report that executes in a reasonable amount of time. Better interoperability of the SAP HANA-based high-speed data warehouse with SAP BusinessObjects tools can also be a big advantage. Instead of creating a query for each report in SAP BusinessObjects, you can finally create generic queries that make entire areas of data available to the reporting platform. Improved data loading performance will allow faster and more frequent refreshing of analytical data. Any administrative work related to data model changes can be done in the new system, in less time with less impact on users. Most importantly, the migration process itself to the new platform does not require interruption of analytical systems. It can be done in several ways, in stages, so that users notice only positive changes and any legacy solutions continue to work, at most better.

Bogdan Bilas, Company Director for Information Technology, SOKOŁÓW
SAP HANA a way to optimize uptime
Three key factors influenced Sokolow’s decision to participate in testing SAP HANA. First, the period of financial reporting to our owners has recently decreased, and as a result, our financial and controlling services have even less time to prepare monthly reports. Another reason is technical. The systematic growth of transactional data directly affects the duration of data synchronization from SAP ERP to SAP BW, which increases the time to make the warehouse available to business users. The third reason – is query performance. The increasingly long time from the moment the report is run until the final results are available is quite a problem for us.
The test results turned out to be quite surprising. A tenfold reduction in query response times is an impressive result. The use of query optimization can certainly give even more spectacular results.
We think that SAP HANA is one of the few solutions that are able to make a very significant difference in the quality of business analysts’ work. In our company, about 90 different types of reports are performed every day for the needs of various organizational units. This is quite a challenge, and in view of this, time to access data is absolutely crucial. We think that all Sokolow S.A. employees who deal with business reporting on a daily basis would see a very significant increase in the comfort of work and a much better use of their time. Which, of course, has a financial dimension for the Company.
Bogdan Bilas, Company Director for Information Technology, SOKOŁÓW
A platform for demanding applications
SAP HANA technology makes it easier for companies to achieve a significant competitive advantage through real-time decision-making. In today’s fiercely competitive market, with a rapidly changing business environment, it is the immediate analysis of available data that is the key to enterprise success. HP, in close cooperation with SAP, has prepared predefined hardware configurations that best meet security and performance requirements. Wanting to prove the usefulness of this innovative technology in business reality, we decided together with BCC to carry out this project. For the tests, performed with Sokolow, we used the HP ProLiant DL 980 platform, dedicated to the most demanding and stressful applications.
When we embarked on the tests, we wanted to see how much benefit SAP HANA technology combined with high-performance HP ProLiant DL 980 devices could bring in practice. Above all, we expected several times faster reporting performance on the tested platform. The test results definitely met our expectations and confirmed that in-memory technology can successfully replace traditional reporting models and technologies. We are convinced that this solution will provide customers with better productivity and help reduce costs.
Huge amounts of data, increasing very dynamically, by up to 50% per year – this is now a problem of every large and increasingly also medium-sized enterprise. To be able to transform this information into the knowledge necessary to make accurate decisions quickly, innovative solutions such as SAP HANA technology are needed. It’s a solution that is sure to work and benefit enterprise-class companies, public institutions, but also many innovative mid-sized companies.
Daniel Slawniak, Alliance/ISV Manager, Hewlett-Packard Poland