Optimizing Imperfect Coverage Cloud-RAID Systems Considering Reliability and Cost

Abstract

This paper considers a cloud provider selection problem to optimize the cloud-redundant array of independent disks (RAID) storage system subject to imperfect coverage (IPC), where an uncovered disk fault causes extensive damages to the entire system despite the presence of adequate redundancy. Given available cloud storage disk providers with disks having different costs, failure parameters and coverage factors, the objective of the optimal design is to select the combination of cloud providers minimizing system unreliability or cost. Both unconstrained and constrained optimization problems are considered. The solution methodology encompasses an analytical, combinatorial method for reliability analysis of the considered cloud-RAID storage system with the IPC behavior. Based on practical design parameters, the brute force approach is utilized to obtain the optimal design configuration. Several case studies are provided to illustrate the proposed optimization problems and solution method.