Research And Application Of Cloud Computing Resource Management System Based On The Open Source Platform

With the fast development of cloud computing, the demand of users for massive computing resources are increasing rapidly, and for the enterprises which provide the cloud computing resources, the way to make the management of the cloud computing resources stable and efficient has become a more and more important problem. On the other hand, although these open source cloud platform softwares have some advantages, such like free of charge,as a result of that, open source platforms attracte more and more attention of the enterprises, but the function of resource management of these open plarforms are not perfect yet, including the automation and the efficient of their management remain to be improved.In order to solve the problems above, in this thesis,according to an enterprise project background, do researches in OpenStack open source platform and Cloud Foundry platform about how to achieve an effective management of cloud resources, and how to make an development of resource management in these two cloud computing platform. Resource management system firstly realizes the virtual machine, real-time monitoring of resources usage, and achieves the automation of resilient and elastic platform resources management. And then, with the help of analyzing the virtual machine monitor and record the use of resources, analyse the amount of resources to predict a period time of the load of virtual machine.According to the prediction made before,develop resource allocation strategies,and make an appropriate and timely expansion or contraction of the resources,so that the stabilization of the virtual machines can be guaranteed, and the platform will be efficient in the use of these platform resources.At the same time, in the open-source cloud platform,the virtual machine that provide the running environment for the application, its performance of the load balancing have the defects,so management system need to optimized for these virtual machines load balancing.By considering the resources consumption characteristics of each application instance,the management will be able to find a way to make a dynamic distribution policy formulation, including make up the error may which caused by network latency incurred in the development of distributed decision-making instances. Finally, resource management system combines server and virtual machine resource monitoring information, according to the system design strategy,with the help of the interfaces the platform provide, realize the function that the automated live migration of virtual machine and enhance the load balancing of the servers at the same time.The objective of this thesis is to provide a stable and efficient resource management on the open source cloud computing platform, achieve the cloud platform to build the resource management system, then accomplish the function of provide a more automated resource management, meanwhile, propose an mechanism for optimizing the distribution of application instances, improve the load balancing status of virtual machines running node. Also with the study of virtual machine migration strategy, by using a comprehensive analysis on the consolidated operating conditions of virtual machines and servers, achieve automate virtual machine live migration. At last, system testing experiments show that the platform management sysytem has meet the design objectives and requirements.