A new approach to improve load balancing for increasing fault tolerance and decreasing energy consumption in cloud computing

Abstract

Listen

Translate article

In recent years cloud computing has been considered among superior technologies worldwide. The main reason is the services and sources provided the users by the clouds. In order to avoid over interactions of the servers and given the work volume and green computations, load balancing in cloud computing is of enormous importance requiring dynamic load distribution in a proportionate manner among the servers. Load balancing may reduce the used energy through avoiding over interaction between the nodes and virtual machines providing desired resource utilization. When a system fails, high costs are imposed on both server and customer, thus load balancing algorithm needs good fault tolerance. There are various techniques to increase fault tolerance. In this study task replication technique was used. To do so three fuzzy inference engines with an approach to fault tolerance for tasks prioritization, virtual prioritization and virtual machines as a goal for task replication were designed. Fuzzy method was selected because the question environment is uncertain and the parameters determination which is carried out by fuzzy method. In the proposed procedure by tasks and virtual priority, we could provide a proper work load distribution. The aim of this study was to present a novel strategy to improve load balancing for increasing fault tolerance and reducing energy consumption via ranking the tasks and virtual machines in cloud computing by fuzzy method.