Abstract

Abstract: File replication is a common strategy to improvedata reliability and availability in large clusters. Reliability for each fileunder server failures based on the relationship between file reliability andreplication factor when servers have a certain probability to fail. In theexisting system more  number of replicacreates it require more energy consumption, time and storage space and some                 time system fails and cannotsend immediate response to the user request. We are implementing energyefficient adaptive file replication system using bloom filtering to reducelatency time. In the propose system increasing number of file replica accordingto user request or user priority and vice versa. Propose strategies in reducingfile read latency, replication time and power consumption in large cluster. Ifmultiple user send request for one file then system create multiple replicaaccording to priority and get immediate response to the user.

Keywords: Data-intensive clusters, file replication, replica placement,energy-efficient