Time Effective Workflow Scheduling using Genetic Algorithm in Cloud Computing

Abstract

Abstract-Cloud computing is service based technology on internet which facilitates users to access plenty of resources on demand from anywhere and anytime in a metered manner i.e. pay per usage without paying much heed to the maintenance and implementation details of application. As cloud technology is evolving day by day it is being confronted by numerous challenges, such as time and cost under deadline constraints. Research work done so far mainly focused on reducing cost as well as execution time. In order to minimize cost and execution time previously existing workflow scheduling model known as predict earliest finish time is used. In this research work we have proposed a new PEFT genetic algorithm approach to further reduce the execution time on this model. A strategy is developed to let GA focus on to optimize chromosomes objective to get best suitable mutated children. After obtaining a feasible solution, the genetic algorithm focuses on optimizing the execution time. Experimental results show that our algorithm can find better solution within lesser time.

M-dimension hybrid algorithm for scientific workflow in cloud computing

“…In [19], writers have presented the novel strategy of PEFT genetic algorithm for decreasing time of execution on this model. One approach is deployed for allowing GA to concentrate on chromosomes aim optimization for getting the best appropriate mutated kids.…”

Task Scheduling in Cloud Computing Environment Using Advanced Phasmatodea Population Evolution Algorithms

“…Nevertheless, the research only considered the makespan under different resource quantities when evaluating. Nagar et al [41] used a previous workflow scheduling model that predicted the earliest completion time, and reduced the execution time by proposing a new Predict Earliest Finish Time (PEFT) genetic algorithm. However, the technique is best suited for workflows with a small number of tasks and does not consider the execution cost, the number of virtual machines, or the data center's energy usage.…”