Design and Implementation of a ROLAP Cube in Scalable Distributed Data Structure

Résumé

The Scalable Distributed Data Structure (SDDS) is a data model specifically designed for distributed environments. An SDDS file comprises records that are dynamically distributed across servers using an SDDS algorithm. A notable feature of SDDS is the removal of a centralized addressing component, simplifying client-server communication and reducing both the message count and data access time in distributed systems. This work also explores a Data Warehouse (DW) within a decision support system, where multidimensional data are represented as a cube and managed through Relational Online Analytical Processing (ROLAP). Although extensive research has been conducted in both the data warehousing and SDDS fields, no prior studies have combined these two areas. This paper introduces a novel approach to implementing a ROLAP cube within an SDDS using the Linear Hashing algorithm (LH*), which eliminates centralized addressing, enabling direct client-server communication and improving performance by reducing inter-site message exchanges. This work demonstrates the feasibility of this method and its positive impact on data processing efficiency in distributed systems.

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Auteur(s)

Statistiques

Consultations : 2

Téléchargements : 0