A Study on Performance of Dominant Scheduling Algorithms on Standard Workflow Systems in Cloud

Abstract

Cloud computing enables users to access computing resources over internet. Organizations that manage and provide these resources are called as cloud service providers. There are several such providers in the market offering large range of services to their users and charge them only for what they have consumed. Workflow is of a set of interdependent tasks that need to be performed in a coordinated way to complete a job. Any business or scientific application can be modelled as a workflow. An effective scheduling algorithm will optimize the resource utilization and in turn the amount spent for executing the workflow. Due to the numerous advantages in cloud, workflow scheduling in cloud seems to be an eminent way to minimize the execution time and cost. There are plenty of existing algorithms and many new algorithms are in the row for workflow scheduling in cloud. A complete understanding of these algorithms will help future researchers to propose novel ideas. Hence this paper provides an analysis of some dominant workflow scheduling algorithms and their performance in standard workflow data sets. The analysis is carried out in WorkflowSim environment with the identical set of Virtual Machines (VM) having similar configuration.