Overview: Semantically Partitioned Objects

A semantically partitioned object (SPO) allows modeling for large data volumes of InfoCubes or DataStore Objects (DSOs). You can use an SPO to more efficiently manage large data volumes to reduce costs. An SPO comprises several InfoCubes or DSOs with data that is partitioned based on certain characteristic values.
To define an SPO, define a reference data model—a standard InfoCube or DSO—for the data target. The BW system creates the criteria for the semantic partitions and the transformation. It also generates the partitioned data targets, the associated transformations, and the data transfer processes (DTPs). Changes are reflected in the partitions, so only the reference data model must be maintained.
These features are required. For more information, see “Using Semantic Partitioning” in the SAP NetWeaver BW 7.3 help.
  • When you select an SPO, the BW Source Designer creates jobs for all SPO partitions.
  • When you select an SPO partition data target, the BW Source Designer creates jobs only for that particular partition.
  • Metadata is read from the SPO reference data structure.
  • Jobs are generated for each partition, including fact tables, star schema, and dimensions.
  • Dimensions are only for those SPOs that are not optimized for SAP HANA.
  • Master data jobs are shared for all SPO partitions.
  • An additional job is created to append the star schema views for partitions into a one-star schema table.