SISYPHUS: A Chunk-Based Storage Manager for OLAP Cubes

In this paper, we present SISYPHUS, a storage manager for data cubes that provides an efficient physical base for performing OLAP operations. On-Line Analytical Processing (OLAP) poses new requirements to the physical storage layer of a database management system. Special characteristics of OLAP cubes such as multidimensionality, hierarchical structure of dimensions, data sparseness, etc., are difficult to handle with ordinary record-oriented storage managers. The SISYPHUS storage manager is based on a chunk-based data model that enables the hierarchical clustering of data with a very low storage cost. Moreover, it provides an access interface that is "hierarchy aware" and thus native to the OLAP data space. This interface can be used to implement efficient access paths to cube data.

• Type: Conference
• Year: 2001
• Where: DMDW
• Authors: Nikos Karayannidis, Timos K. Sellis
• Keywords: DMDW 2001, Management, OLAP, SISYPHUS Storage Manager, Storage Manager