Workflow scheduling with data transfer optimisation and enhancement of reliability in cloud data centres

Abstract

Infrastructure as a service (IaaS) clouds offer huge opportunities to solve large-scale scientific problems. Executing workflows in such environments can be expensive in time if not scheduled rightly. Although scheduling workflows in the cloud is widely studied, most approaches focus on two user's quality of service requirements namely makespan (i.e., completion time) and costs. Other important features of cloud computing such as the heterogeneity of resources and reliability must be considered. In this paper, we present a reliability-aware method based on discrete particle swarm optimisation (RDPSO) for workflow scheduling in multiple and heterogeneous cloud data centres. Our aim is to optimise data transfer time while minimising makespan and enhancing reliability. Based on simulation, our results show a significant improvement in terms of makespan, transferred data and reliability relative to reliability-aware HEFT method (heterogeneous earliest finish time), for the real-world workflows.