An Evaluation of Data Consistency Models in Geo-Replicated Cloud Storage

Cloud computing has become a major and most preferred technology for data storage across the world. This technology tremendously improved service delivery to clients as data and services are brought closer to the client. Bringing data and services closer to the client requires the implementation of a distribution architecture, that is, an architecture that spans across a wider geographical area. However, such architecture comes with its own challenges. It requires proper management of data and services across multiple data centres to maintain consistency, availability of data and services, and systems that are tolerant to failures.. In this study we focus on evaluating currently used methodologies to ensure data consistency performance levels for cloud storage and determine whether the proposed solutions are effective and feasible. From the surveyed data consistency models, we noted that it's difficult to find one model that satisfies the three main challenges associated with data consistency across multiple geo-replicated data centres.