Skip to Content

FAQ - BI Accelerator - Technical Aspects - Query Performance

What makes SAP NetWeaver BI Accelerator so fast?

Traditional approaches to the optimization of query response times require some form of pre-aggregation. SAP NetWeaver BI Accelerator performs aggregation quickly during query execution because it processes the entire query in memory with no disk reads. 

The BI Accelerator owes its speed to a series of breakthrough innovations; for some, patents are pending. Query processing is performed entirely in memory, using extremely compact index structures that are persisted on a separate file server with gigabit connectivity to avoid the need for database reads. To create the indexes, database tables are decomposed vertically and stored by column, not row, and the indexes can be split horizontally for highly parallel processing on multiple processors. Smart compression is used based on dictionaries and integer processing, and optimized to run on 64-bit Intel CPUs. The algorithms are innovative and use TREX search engine technology. Flexibility is increased by adopting the adaptive computing paradigm. SAP NetWeaver BI Accelerator runs on blade server infrastructure that allows for a dynamic assignment of hardware resources to ensure high availability and good load balancing.

How is SAP NetWeaver BI Accelerator integrated into BI functionalities?

SAP NetWeaver BI Accelerator can be used in (almost) all cases where aggregates can be used:

  • Real-time InfoCubes: closed requests can be indexed by the SAP NetWeaver BI Accelerator, open requests will be read from database 
  • MultiProviders: any indexed InfoCube as part of the MultiProvider will use SAP NetWeaver BI Accelerator. Note that overall query performance is defined by the slowest InfoCube in a MultiProvider environment 
  • Inventory handling: InfoCubes containing non-cumulative key figures can be indexed

Note: there is only one expceptional scenario where SAP NetWeaver BI Accelerator and aggregates are not congruent: With InfoSets, an SQL statement is generated and, hence, the SAP NetWeaver BI Accelerator cannot be used.

Is SAP NetWeaver BI Accelerator used for value help?

If the Query Execution Filter Value Selection is set to Q / Only posted value for navigation for a characteristic, the value help in query execution uses the data manager. This means it also uses the BI Accelerator (similar to how it uses aggregates) if applicable (for retrieving the SIDs, texts are read from the database).

Which interfaces use SAP NetWeaver BI Accelerator?

SAP NetWeaver BI Accelerator is used for all queries on InfoCubes which are indexed. In addition to Business Explorer, the following features/interfaces also make use out of SAP NetWeaver BI Accelerator:

  1. OLAP interfaces: XMLA, ODBO, OLAP BAPI 
  2. listcube 
  3. custom-specific implementation based on INFOPROV_READ

How can I find out if SAP NetWeaver BI Accelerator helps to solve performance issues?

Please check the statistics for all critical queries (e.g. via Query Monitor RSRT or via technical content queries/Admin Cockpit). Just have a closer look at the share of database time (opposed to OLAP or frontend time); SAP NetWeaver BI Accelerator may cut down the database time almost arbitrarily.

Is it recommended to switch off the OLAP cache once SAP NetWeaver BI Accelerator is used?

There is no reason to switch off the OLAP cache, and we recommend that customers keep the OLAP cache switched on for InfoCubes with BI Accelerator indexes. The cache settings should be the same as for InfoCubes with aggregates. Using OLAP Cache takes load off SAP NetWeaver BI Accelerator and contributes to its high scalability.

Former Member

No comments