Research On Dynamic Optimization Of Cloud Computing Performance And Energy Saving

With the progress and development of Internet information technology,it is an important research direction to find a computing method of high performance and low cost in the field of computer science.At the same time,along with the development of distributed computing,grid computing,utility computing technology,the researchers compromise the advantages of these traditional computer science and network technology,and present a more advanced computing method--cloud computing,which is a great change in the field of the Internet.Cloud computing develops rapidly by then.However,with the rapid development,the problem of low efficiency of task scheduling and the high energy consumption of datacenters is also coming.Therefore,in this paper,we study the model of datacenter and the task scheduling strategy to improve the performance of cloud computing.We have studied the performance evaluation and optimization of systems and Petri net for a long time,and there are many researches about modeling of Petri net.Therefore,combining the key components of cloud computing and the corresponding task scheduling,a generalized Stochastic Petri Nets based on the task scheduling is presented.Based on the model,we design the strategy of scheduling based on the genetic algorithm to improve the efficiency and reduce the energy consumption,and we carry out the experiments to verify the effectiveness of the strategy.The main contents of this paper are as follows:1.Research and establishment of cloud computing task scheduling process model.This paper studies the current cloud computing data center structure,summarizes the main components of the cloud computing data center(server,virtual machine,etc.),analyzes the general process of cloud computing task scheduling processing;the corresponding elements and events in the process are presented by Petri net,the cloud computing task scheduling processing refinement model is established.The model is simplified by Petri net model equivalence and simplification,and the properties of the model are simplified and the correctness and feasibility of the simplified model are verified.2.Research and designing of cloud computing task scheduling strategy.Based on the task scheduling process model designed in this paper,the genetic algorithm is selected used to find the entry point of the combination of genetic algorithm and task scheduling process.It is integrated into the dynamic voltage scaling mechanism of the server and the task scheduling strategy based on genetic algorithm is designed.The dynamic task scheduling problem is solved according to the minimum of the total task completion time and the minimum energy consumption required for the task completion,and further the experimental analysis is carried out.After the theoretical analysis,model construction,strategy designing and implementation,the final experimental results shows that the proposed virtual machine-based task scheduling strategy can shorten the total time of task completion and reduce the total energy consumption of all tasks.The validity of the task scheduling strategy proposed by the topic.