Quality-oriented Copy Placement And Update Algorithms

With the development of networks, network data information is growing rapidly. In order to make the information which distributed in the widely geography to share the resources, the replica technology is proposed by people. Replica technology can effectively improve the access efficiency of data processing, reduce data access latency, also can ensure the network load balancing and data reliability, security and fault tolerance.Replica technology mainly involves three problems:1) when the replica is created, 2) what kind of the filereplica is created,3) and how to place replicas. Among them, replica placement is the key problem of replica technology. The replica placement problem is how to properly placed replicas, and how many replicas should be placed in networks. Efficient approaches and strategies for replica placement play a crucial role not only in improving availability, reliability of the systems, but also in reducing the user access time and improve the access efficiency of data processing.In order to ensure the quality of service and reduce the cost of network operation, we propose a replica placement algorithm, Multiple Big QoS Far Servers First. Experimental results show that, the proposed algorithm successfully maximizes the processing capacity for each replica, while significantly reducing the running cost of the network under the different distance limits. The best improvement is up to 68%in comparison to the state-of-the-art.Faced with the dynamic changes of data access requirements in the network, we propose three replica update algorithms in different strategies 1) GrRP algorithm,2) TS (tabu search algorithm),3) Min Placement Far Servers First. In time complexity, the execution time is reduced up to 87.9%, and the cost of network operating and replica update is no more than 2.5% in comparison to the existing dynamic programming algorithm. In cost, Experimental results show that, we proposed the algorithm with update policy which compared with existing algorithm for replica placement, successfully maximizes the processing capacity for each replica and improve the access efficiency of data processing, reduce data access latency, the running cost of the network gets a considerable reduction under the different distance limits, and the best improvement is up to 84.6% in comparison to the state-of-the-art.