New Fuzzy-Based Fault Tolerance Evaluation Framework for Cloud Computing

Abstract

Fault tolerance is one of the principal challenges in cloud computing. This capability has a trade off with other system features. Providing a fuzzy inference system to evaluate fault tolerance architectural capabilities in cloud computing systems is among the goals of this research. The proposed system is called CFTFIS. It evaluates the fault tolerance architecture and determines the level of fault tolerance based on fuzzy patterns. Three parameters of each architecture, i.e., policies, fault detection and fault recovery techniques, have been considered. Fault tolerance of the presented architecture and the increased percentage of the intended capability are obtained at five different fuzzy levels. If the vulnerability of a cloud computing system is identified, it is possible to make a correct and proper choice of the architecture that should be implemented in the system. Additionally, if the damage to a cloud computing system is defined based on the amount and type of the faults, it is possible to choose an architecture and implement it in the system based on the fault tolerance that each architecture provides.