Abstract. Data Replication Broker is one of the most important components in data grid architecture as it reduces latencies related to file access and file transfers (replica). Thus it enhances performance since it avoids single site congestion by the numerous requesters. To facilitate access and transfer of the data sets, replicas of data are distributed across multiple sites. The effectiveness of a replica selection strategy in data replication broker depends on its ability to serve the requirement posed by the users jobs or grid application. Most jobs are required to be executed at a specific execution time.

To achieve the QoS perceived by the users, response time metrics should take into account a replica selection strategy. Total execution time needs to factor latencies due to network transfer rates and latencies due to search and location. Network resources affect the speed of moving the required data and searching methods can reduce scope for replica selection. In this paper we propose an approach that extends the data replication broker with policies that factor in user quality of service by reducing time costs when transferring data. The extended broker uses a replica selection strategy called Efficient Set Technique (EST) that adapts its criteria dynamically so as to best approximate application providers’ and clients’ requirements. A realistic model of the data grid was created to simulate and explore the performance of the proposed model. The policy displayed an effective means of improving the performance of the network traffic and is indicated by the improvement of speed and cost of transfers by brokers.