An asynchronous replica consistency model in Data Grid

Abstract

Data Grid is an emerging technology to provide uniform access and management of the large scale distributed scientific datasets. To enhance the accessing performance, replica is an ideal solution for the Data Grid. But in the wide-area network environment, to maintain the consistency of those replicas is not a simple task because of the resource's autonomy and the network's instability. In this paper, a new asynchronous model is proposed, which could avoid the inconsistency in spite of the possible network congestion and system failure. This model could be more suitable for the resource autonomy characteristic, putting fewer burdens on storage resources, and it is also easier to implement while providing more scalability for the grid system. And then a detailed implementation, use case and some further research plans are described. © Springer-Verlag Berlin Heidelberg 2005.