CCESHP: Causal Consistency Model of Edge Storage Based on Hash Ring and Partial Geo-Replication

Abstract

Abstract At present, most of the causal consistency models that rely on cloud storage have problems such as high operation delays and large metadata overhead. To solve these problems, this paper proposes a causal consistency model for edge storage based on hash rings, CCESHP. The proposed model uses two hashes to map the keys and servers on the hash ring for grouping and stores a subset of the complete data set in a replica node located at the edge of the network, thereby realizing a partial geographic replication strategy in the edge storage environment. Operation latency will be reduced since the edge replica is closer to the client. At the same time, it also generates and maintains a combined timestamp to capture causality according to the update type, which can keep the amount of managed metadata in a relatively stable and low state, reduce the overhead of system management metadata, and improve system throughput. The experimental evaluation results under different workloads show that the model has better performance in throughput and operation delay when compared with the existing causal consistency model.