Settings for Extraction and Data Transfer

For requests that end in the PSA: Data Transfer Using qRFC Outbound Queue

For source systems with Service API release 7.0 or higher (Plug-In Basis 2005.1) or for 4.6C source systems (Plug-In 2004.1 SP10), requests that end in the PSA use qRFC with outbound queues by default. Therefore they use the QOUT scheduler (transaction SMQS) for data transfer. These are PSA requests in the new data flow with DTP and PSA requests for 3.x DataSources that do not write to the data targets (requests for 3.x DataSources with the InfoPackage processing option Only PSA).

In the QOUT Scheduler, specify the maximum number of dialog work processes used for the data transfer and the maximum processing time of the scheduler for the queue.

More information:

Data transfer with qRFC outbound queue

SAP Note 1163359

For requests that do not end in the PSA: SAPI-controlled data transfer

For source systems with Service API release 7.0 or higher (Plug-In Basis 2005.1) or for 4.6C source systems (Plug-In 2004.1 SP10) and for requests that do not end in the PSA, the Service API is used to control the data transfer. These are PSA requests for 3.x DataSources that also write to data targets (for example, requests for 3.x DataSources with the InfoPackage processing option PSA and then packages to data targets) and requests in a data mart scenario (for example, updating a DataStore object in an InfoCube using an Export DataSource). You specify the control parameters for data transfer for all the DataSources of a source system by using the transaction SBIW (in the source system) under General Settings Maintain Control Parameters for Data Transfer . The parameters are saved in the table ROIDOCPRMS. In BW you can overrride the DataSource-specific part of the settings in the InfoPackage under  Scheduler Default DataSource Settings for Data Transfer . The parameters are saved in the table ROOSPRMS.

The main memory requirement for extraction depends mainly on the size of the data package and can become very large. The package size depends on various parameters. You can reduce the memory consumption by optimally configuring the settings for control parameters. For more information and examples about specifying the parameters that influence the data package size, see SAP Notes 417307 and 409641.

In addition, you can influence the performance by optimally setting the maximum number of dialog processes for parallel processing. You can then specify whether a particular application server should be used for the background job for extraction or whether the system should choose the most appropriate application server using load balancing.

More information:

Control parameters in the source system: Maintaining Control Parameters for Data Transfer

DataSource-specific override of control parameters in the BW system: Maintaining Data Transfer Control Parameters for DataSources

Enhancing DataSources

If you enhance the DataSources and fill the fields of the append structure using Exit or BAdl, note the information under Performance Tips for ABAP Programming. If customer enhancements are very time consuming (for example, when using ABAP runtime analysis or an SQL trace), check and improve your coding.