Fault tolerant aware energy-efficient VM integration in Cloud Computing

Abstract

Abstract Cloud computing has turned the IT industry by providing a service-oriented approach to computing. Users can access cloud computing services through a simple portal, even if they are not familiar with the underlying systems. Cloud computing systems perform many complex and challenging operations in addition to managing a large infrastructure of requests. These complex processes pose a number of challenges for service providers including reliability, security, sustainability, energy consumption, and management of the concern resources. This paper focuses on two of the most important challenges facing cloud computing providers: reliable resource provisioning and energy efficiency. These two challenges are often contradictory, as reliable resources often consume more energy. This paper proposes a new resource management policy that aims to deploy the most reliable and energy-efficient resources and allocate them to the appropriate virtual machines. The paper also proposes a mathematical framework for modelling the relationship between energy consumption and reliability. The proposed resource management policy significantly reduces service disruption or downtime, compared to environments that assume independent failure events. The policy also improves the reliability-to-power consumption ratio by 14%, according to the mathematical model.