Sap Ag

Christian M Bartholomae

Guenter Radestock

Franz Faerber

Stefan Dipper

Erich Marschall

Wolfgang Stephan

Klaus Nagel

Thomas Zurek

Stefan Unnebrink

Franz-X Faerber

A query having multiple parts may be processed to form an intermediate results set. This intermediate results set may be partitioned into a plurality of groups. Thereafter, the groups may be sorted into a plurality of containers so that each container contains data sufficient to calculate one requested result in the multipart query. Related techniques, apparatuses, systems, and computer program products are also described.

Efficient handling of multipart queries against relational data

A first data storage schema in which a characteristic in a first dimension table is mapped by a first table and a second table can be received and a second data storage schema can be received. The first table maps the characteristic to a first object that include attributes to which time information is irrelevant to data processing activities and the second maps the characteristic to a second object that includes attributes to which time information is relevant to data processing activities. The second data storage schema includes a fact table including at least some facts drawn from the first data storage schema and a second dimension table that includes at least some characteristics drawn from at least one of the first object and the second object.

Mapping data structures

A method is disclosed for modeling application-level objects in terms of join graphs defined over tables containing structured data residing in a relational database. In accordance with the disclosed metamodel, each object is modeled logically as a join graph. A query received from an application that requests the return of objects meeting certain conditions is reformulated to refer to the metamodel. The metamodel includes an index structure having a plurality of indexes and a set of join conditions that specify relationships between the indexes. Some series of join conditions form join paths, such that each join path originates on an anchor table and ends on a table corresponding to one of the plurality of indexes. The metamodel further includes at least one view representing a subgraph of the join graph having at least one anchor table as a key.

Object metamodel comprising views on a join graph

A system and method for executing a query plan are disclosed. In the system and method, a join graph is generated to represent the query plan. The join graph includes a set of plan operations that are to be executed for implementing the join graph. The query plan is received by a distributed network of a logical index server and one or more selected physical index servers. Each physical index server receives a portion of the plan operations, and determines what plan data is needed to execute the portion of the plan operations. A system and method includes a process for determining what plan data is needed from other physical index servers, or what plan data is needed by other physical index servers.

Growing community of inventors

Oftersheim, Germany

Christian M Bartholomae