View divergence of freshness in replicated database systems: A review

Abstract

Data replication is often used in distributed database applications to improve database availability and performance. Replicated database must be periodically refreshed using update propagation strategies. The concept of freshness is used to measure the deviation between replica copies. In this paper, we analyze the different methods to control the view divergence of data freshness for clients in replicated database systems whose facilitating or administrative roles are equal. The methods, mentioned in this paper, provide data to clients with freshness specified by them when updates are initially accepted by any of the replicas, and then, asynchronously propagated among the replicas. To provide data with statistically defined freshness, we select multiple replicas using a distributed algorithm and centralized algorithm so that they receive all updates issued up to a specified time before the present time. We compared the view divergence control of both methods in terms of controlled data freshness, time, message, and computation complexity. The comparison showed that distributed method achieves a great improvement in data freshness compared with centralized method.