Article ID Journal Published Year Pages File Type
4956557 Journal of Systems and Software 2016 22 Pages PDF
Abstract
Hybrid cloud storage combines cost-effective but inflexible private storage along with flexible but premium-priced public cloud storage. As a form of concurrent sourcing, it offers flexibility and cost benefits to organizations by allowing them to operate at a cost-optimal scale and scope under demand volume uncertainty. However, the extant literature offers limited analytical insight into the effect that the non-stationarity (i.e., variability) and non-determinism (i.e., uncertainty) of the demand volume - in other words, the demand variation - have on the cost-efficient mix of internal and external sourcing. In this paper, we focus on the reassessment interval - that is, the interval at which the organization re-assesses its storage needs and acquires additional resources -, as well as on the impacts it has on the optimal mix of sourcing. We introduce an analytical cost model that captures the compound effect of the reassessment interval and volume variation on the cost-efficiency of hybrid cloud storage. The model is analytically investigated and empirically evaluated in simulation studies reflecting real-life scenarios. The results confirm that shortening the reassessment interval allows volume variability to be reduced, yielding a reduction of the overall costs. The overall costs are further reduced if, by shortening the interval, the demand uncertainty is also reduced.
Keywords
Related Topics
Physical Sciences and Engineering Computer Science Computer Networks and Communications
Authors
, , ,