Research And Implementation Of Fair Allocation Algorithm For Maximizing Virtual Resource Utilization In Cloud Platform

As cloud computing technology matures,more and more companies or individuals tend to submit computing tasks or storage tasks to the cloud platform for execution.Therefore,a reasonable allocation of resources in the cloud platform can effectively improve the efficiency of computing or storage tasks.At present,the resource allocation methods in the cloud platform are mainly the first input first output(FIFO)and the proportional fairness(PF).However,these two methods will cause uneven resource allocation and waste of resource,resulting in lower resource utilization.For improving the utilization of resource,the dominant resource fairness(DRF)algorithm was proposed to satisfy the definition of fairness better than the traditional resource allocation method in cloud platform.Although DRF achieved an higher resource utilization than FIFO and PF,it still has room for improvement in resource utilization.This paper proposes a fair distribution mechanism based on weighted virtual prices,which gives different virtual prices to each resource,and ensures that the minimum resource utilization is maximized on the premise that each user's virtual assets are equal.At the same time,a fast algorithm is proposed by analyzing the combined nature of the mechanism.The comparative experiments are conducted to compare the resource utilization of the proposed mechanisms against the existing mechanisms,including the DRF mechanism,asset fairness(AF),and competitive equilibrium from equal incomes(CEEI).The experimental results show that the fair distribution mechanism can effectively improve the resource utilization rate of cloud computing platform.