Research On Resource Management Technology In Cloud Computing

In recent years,the word Cloud Computing is getting more and more popular in people's life.Cloud computing is a kind of based on the calculation of the Internet,in this way,the sharing of software and hardware resources and information can be according to the demand for computers and other equipment.At present,the concept of "cloud computing" has been used in actual production environment.Cloud Computing is expanding its services and influences.Cloud computing service need to rely on data center eventually.With the development of the Internet and mobile Internet,the role of data center is also increasing.There are more and more companies choose to build cloud data center,using the cloud computing platform management technology to maintain a large number of servers to work together,the realization of the automatic and intelligent large-scale system reliable operation.This leads to a number of data centers and are growing rapidly.In the process of data center development,the deficiency in the data center of resource management,lead to the phenomenon of low utilization of more and more get the attention of people,some of the data center resources utilization even less than 20%on average,which led to a huge waste of resources.Another problem is that the energy consumption,due to the resource utilization is low,many servers without load conditions lead to energy drain,which lead to the data center costs high.This passage embarks from the cloud data center resources management deficiencies,based on resource management technology in the most popular open source cloud management platform OpenStack,especially the virtual machine scheduling policy,researching its existing scheduling mechanism and the working process.Aiming at the existence of its defects designed and implemented a new solution,which based on improved ant colony algorithm it virtual machine deployment strategy.This strategy takes advantage of the ant looking for food in the process of distributed collaboration and feedback mechanism to deal with the platform is the virtual machine batch create request.On the platform of network resources management problems,the lack of network traffic monitoring function in this paper,on the basis of it the existing network architecture,by increasing the OpenFlow controller,realize the port traffic statistics for each virtual machine.Then virtual machine instance is obtained by computing the real-time traffic information.Finally through constructing real experimental environment to verify the improvement of the availability,reliability and scalability of the platform,and access to a larger virtual machine request simulation method is simulated,the performance of the improved ant colony algorithm deployment strategy for validation,proved that the strategy can be improved in both energy consumption and performance under the premise of optimal deployment of results are obtained.The virtual machine network traffic real-time monitoring function through the contrast test,verify the availability and reliability of the network flow monitor.