Time-Cost Effective Algorithms for Cloud Workflow Scheduling - Extension of An Earlier Work

摘要

This paper is an extension of an earlier work carried out by Babu et al [22]. Two scheduling algorithms are presented in this paper, namely, Time-Constrained Early-Deadline Cost-Effective Algorithm (TECA) to schedule these time critical workflows with minimum cost and, Versatile Time-Cost Algorithm (VTCA) which consider both time and cost constraints. TECA schedules tasks to complete in the earliest possible time and optimizes the costs in resource provisioning. VTCA supports Quality of Service (QoS)-based scheduling keeping a balance between completion time and cost for the selected QoS level. Both algorithms schedule tasks of same height within the minimum completion time (using Max-Min algorithm). The tasks get scheduled on new resources only when their completion times are more than the calculated minimum completion times for the given resource. CloudSim-based results show that our algorithms minimize completion time better than other popular algorithms, in addition to reducing costs. The modeling for costs satisfies the criteria of earliest completion time.