Energy-aware workflow scheduling algorithm for the deployment of scientific workflows in cloud

Abstract

Recently, there is a fast increase in the use of cloud computing and its assets for the execution of scientific workflows [1] (Xu et al. IEEE Trans Cloud Comput 4(2);166–179, 2016). All the scientific applications are very broad in scale, so the fruitful execution of the application can be in fact finished by expanding in a cloud platform. A single scientific workflow more often contains a huge number of tasks; to execute all these tasks, it requires more number of resources. Only cloud infrastructure can provide such a huge number of computing resources. All the processing assets are given as virtual machines inside a cloud platform. More amount of computational energy is spent to execute complex scientific applications, so it becomes very critical to use the cloud virtual machines in an energy conservational way. So, using the cloud computing resources in more energy efficient manner has gained more attention among the analyst. But even today, executing a scientific workflow in an energy-aware manner in the cloud platforms is a challenging task. Considering all these issues, this paper proposes an energy-aware workflow scheduling with task migration (EAWSTM) algorithm which is proposed for the deployment of various scientific workflows in cloud platforms.