Abstract

Abstract:File replication is a common strategy to improve data reliability andavailability in large clusters. Reliability for each file under server failuresbased on the relationship between file reliability and replication factor whenservers have a certain probability to fail. In the existing system more numberof replica creates it require more energy consumption, time and storage spaceand some time system fails and cannot send immediate response to the userrequest. In the proposed system energy efficient adaptive file replicationsystem using bloom filtering to reduce latency time. In the propose systemincreasing number of file replica according to user request or user priorityand vice versa. Propose strategies in reducing file read latency, replicationtime and power consumption in large cluster. If multiple user send request forone file then system create multiple replica according to priority and getimmediate response to the user.