Techniques for managing what-if analysis of data managed by a relational database system

Abstract

An analytical workspace includes a set of data on which what-if analysis is performed, which is stored persistently and managed in a relational database system. The data may be stored as a binary large object in relational tables, and may comprise a portion of a multidimensional cube. A series of commands on the analytical workspace is received by a database server. In response to the commands, changes are made to the data in temporary storage. Redo information is not generated at this time in association with such changes. Hence, these changes are not made permanent at this time. At some point, the server may receive an update command to make the changes permanent. In response to the update command, redo information is generated for the changes reflected in the temporary storage and the changes are made permanent, in accordance with relational database properties.

Dependencies are inferred from the claim text.