A hybrid approach to replica management in data grids

Abstract

Data grids are current solutions suggested to meet the needs of large-scale systems. They provide a set of various resources which are geographically distributed. A data grid allows quick and efficient access to data, enhancement of availability, and fault tolerance. In large-scale systems, these advantages are possible only by the use of replication. The latter, however, poses the problem of the maintenance of the consistency of replicas of the same data set. In this paper, we propose a hybrid method combining the pessimistic and optimistic approaches conceived for data consistency maintenance. Our approach is based on a hierarchical representation model with two levels, which facilitates the management of replica consistency in the large-scale systems.