Modeling Cloud Storage: A Proposed Solution to Optimize Planning for and Managing Storage as a Service

Abstract

Cloud-computing service providers are currently viewed as the best solution to the global need for massive data systems because of their superior flexibility, scalability, and cost benefits. Cloud computing that is enabled by virtualized services is still constrained, however, by the capacities of the underlying physical systems that are combined into sharable pools of resources. The next challenge for computation systems will arise when even the cloud is not sufficient. What comes after cloud migration and adoption? In this paper, we examine how service providers can manage cloud storage resources and costs when the amount of collected data to be retained grows exponentially, to the point that it strains even virtualized resource capacities. We assess the analytical frameworks being developed to identify which storage architectures can best accommodate the specific needs of large data storage consumers. We also investigate the areas in which these fail to fully address the problem, and propose solutions. We argue that a cloud storage framework that addresses data volume, data growth trends over time, and requirements for storage management will enable service providers to manage cloud storage resources and costs in such a manner that the cloud will continue to offer the greatest benefits for the storage of massive data systems.