Component Importance Analysis of Virtualized System

This paper presents component importance analysis of virtualized system design. The component importance analysis is significant to develop a trusted system in its design phase. This paper discusses the importance of components from the viewpoint of availability. Specifically, based on the fault tree and continuous-time Markov chain models for virtualized system in Kim et al. (2009), we present a new method to evaluate component importance. In numerical examples, we illustrate quantitative importance analysis and compare the availabilities of non-virtualized and virtualized system designs.