Implementation and Testing of Failure Recovery Based on Backup Resource Sharing Model for Distributed Cloud Computing System

Abstract

A cost-effective backup resource allocation model for cloud computing systems based on a probabilistic protection guarantee was presented. This paper implements the backup resource allocation model and tests its failure recovery performance. A prototype cloud system is implemented by using the existing capabilities of OpenStack framework. We observe that a failed virtual machine is successfully recovered by using a shared backup machine in single- and multi-site environments.