Distributed Database Management Systems and the Data Grid

Heinz Stockinger

Digest of papers - IEEE Symposium on Mass Storage Systems · 2001 · 72 citations · 6 references

Concepts

TL;DR

Grid research and distributed database research both address data replication from distinct perspectives, and the target object‑oriented DB, Objectivity/DB, is widely used in current and future CERN high‑energy physics experiments. The paper aims to outline both approaches and identify commonalities to design a most efficient Data Grid for managing data in object‑oriented databases. The authors describe Data Grid characteristics and needs for HEP, base their discussion on the Globus toolkit middleware, and outline both grid and distributed database approaches to identify commonalities for an efficient Data Grid managing object‑oriented database data.

Abstract

Currently, Grid research as well as distributed database research deals with data replication but both tackle the problem from different points of view. The aim of this paper is to outline both approaches and try to find commonalities between the two worlds in order to have a most efficient Data Grid that manages data stored in objectoriented databases. Our target object-oriented database management system is Objectivity/DB which is currently the database of choice in some existing High Energy Physics (HEP) experiments as well as in next generation experiments at CERN. The characteristics of Data Grids are described, especially within the High Energy Physics community, and needs for Data Grids are defined. The Globus toolkit is the Grid middle-ware on which we base our discussions on Grid research.

References

6