Fast replica recovery and adaptive consistency preservation for edge cloud system

Abstract

Edge cloud extends the power of cloud computing to the edge of the devices that are closest to the demands of big connection, low latency and large bandwidth. However, there are still many challenges due to the dynamic, heterogeneous and real-time bandwidth of the node in the edge cloud environment. This paper studies dynamic replica creation, fast replica recovery and adaptive consistency preservation. Before solving the problem of data inconsistency caused by frequent updates of replicas, we consider the strategy of fast replica recovery. Firstly, the DRC-DS is based on the regional structure, considering the number of replicas and the creation location of the data process and copying the data information with high access frequency and long average response time. Moreover, the FRR-LB is based on the heat ranking of the replica, the source node and the target node are determined according to the double-cycle lookup structure. Besides, the ACP-IMP is proposed to solve the problem of replica consistency in the environment of high failure rate of network and node. The more certainty replica nodes are elected as the leaders of edge nodes, and the message transmission of reconfirmation is reduced.