Open Replication of Databases

Informations projet

GORDA

N° de convention de subvention: 004758

Projet clôturé

Date de début 1 Octobre 2004

Date de fin 31 Mars 2008

Financé au titre de

• Information Society Technologies: thematic priority under the specific programme "Integrating and strengthening the European research area" (2002-2006).

Coût total € 1 729 432,00

Contribution de l’UE € 1 218 643,00

Coordonné par

UNIVERSIDADE DO MINHO
Portugal

Résultats exploitables

Database replication as a means to address the challenges of trust, integration, performance, and cost

The goal of the GORDA project is to foster database replication as a means to address the challenges of trust, integration, performance, and cost in current database systems underlying the information society. This is to be achieved by: - Architecture and Interfaces Promoting the interoperability of DBMSs and replication protocols by defining generic architecture and interfaces that can be standardized. - Innovative Replication Providing general-purpose and widely applicable database replication protocols based on group communication for large-scale database systems. - Management Tools Providing uniform techniques and tools for managing secure and heterogeneous replicated database systems. This means supporting, at architectural and protocol levels, complex environments by addressing partial replication, distributed execution and cross-border security issues. - Open Standard Making all interfaces open and actively working to foster adoption of the interfaces and of the resulting generic replication middleware by different database and middleware providers. - Migration Path Providing an interim middleware-based solution that allows immediate integration of current DBMSs, thus supporting the standardisation effort. In its first year the project concentrated on the definition of the GORDA Architecture and APIs (GAPI). The GAPI consists of a generic architecture and interfaces enabling to accommodate different database engines and replication strategies, and offering a set of generic and expressive programming interfaces easing the efficient interaction of the intervening system components. A first proposal for the GAPI has been documented in detail and is being disseminated to the appropriate forums for discussion. Scientific results with the proposal and evaluation of database replication protocols, group communication protocols, and autonomic system management have been published. In the second year of the project the GORDA architecture and programming interfaces (GAPI) have been consolidated and mapped into two major database management systems. Several database replication protocols have been implemented making use of the GAPI and two novel protocols have been proposed. A major effort has been put in the refinement of a simulation infrastructure for evaluation and benchmarking of both communication and database replication protocols, in the implementation of publicly available prototypes of the replication protocols and in the development of one of the industrial partners main commercial product, Sequoia. In the last year of the project, substantial effort was put into the dissemination and adoption of the GORDA architecture and programming interfaces (GAPI). A novel replication protocol, using the GAPI in wide-area network has been proposed. As for group communication, the consortium proposed the GORDA Group Communication Service Specification (GCS), as a draft standard and promoted its dissemination and adoption. A major effort, during this period, has been put in the integration of the work produced by the consortium, which resulted in a demonstration of a running prototype of several GORDA replicated databases.