Research On Energy Consumption And Cost Scheduling Strategy Of DAG Application In Cloud Compution

With the rapid development of information technology in the 21st century,in order to meet the great demand for computing resources in the era of big data,cloud computing emerges at the historic moment.Cloud computing mode can make full use of computing resources as much as possible,and forms an operation mode like water and electricity.However,on the one hand,energy consumption is exploding in today's data centers,supercomputers,and various cloud systems.Green computing is a new trend in industrial development,and a lot of effort and work is focused on task scheduling to reduce energy consumption.In order to make the task scheduling in within the time limit for greater energy saving effect,while maintaining quality of service,based on DAG application in the cloud system scheduling,this paper proposes an adaptive cuckoo search algorithm,which uses a new parameter setting strategy,to minimize the energy consumption of tasks with any priority constraints in the cloud data center.On the other hand,with the continuous expansion of the scale of application programs,the demand for processor processing power is getting higher and higher,and the required processor hardware costs are also increasing.However,the correct execution of programs needs to meet the functional security requirements,including real-time requirements and reliability requirements.Based on this,this paper provides a way to meet the function security of distributed heterogeneous parallel application cost optimization method.The method continuous selectly add processor which has optimal costeffective.Furthermore,under the condition of meeting the deadline requirement,the method constantly choose replica the task which could improve the reliability best,in a relatively short period of time.At the same time,the processor hardware cost is rapidly optimized.The main work is as follows:1.The distributed heterogeneous processors for the cloud data center energy consumption optimization problem,this paper proposes a strategy based on Monte Carlo parameters evaluation of feedback control of adaptive cuckoo search algorithm,the algorithm uses a new encoding mechanism to deal with the priority constraints between tasks and the frequency constraints of the processor so as to minimize the energy consumption of parallel tasks under any priority constraints in heterogeneous distributed heterogeneous systems.At the same time,we use HEFT algorithm to construct a relatively high-quality initial solution.2.For the first time,a parameter feedback control scheme based on Monte Carlo strategy evaluation is used to balance global and local search,in which way its search ability is greatly enhanced,reduces the number of population iteration,and improves the quality of solution.In the end,the proposed selfadaptive cuckoo search approach is validated with two benchmarks and a randomly generated case,and the experimental results demonstrate that our proposed approach outperform some other state-of-the-art approaches.3.Aiming at the hardware cost optimization problem of distributed heterogeneous embedded processors in cloud data center,a novel hardware cost scheduling optimization method is proposed which has the requirement of functional security guarantee based parallel applications.The deadline for application execution,and reliability constraints,on the basis of HEFT algorithm,when scheduling time does not satisfy the constraints,the way choice"cost-effective" highest processor increase processor,this step is repeated until the functional safety requirements are met,and the final processor cost is obtained.The experimental results show that the cost optimization method proposed in this paper is superior to similar methods under the same computation amount,and has a lower algorithm time complexity.