Research And Optimization Of Virtual Machine Deployment Mechanism In Open-source Cloud Computing Platform Openstack

In recently years, the word Cloud Computing came into our field of vision more than before, hence caused more attention which is paid to it.Cloud Computing describes a new model through which IT service is increased, used and delivered based on the Internet. It is always related to provide resources which is dynamic, extensible and usually virtualized. Cloud Computing not only changes the technology, but also changes the business model. In the long run, with the change of business model, whether enterprise users or individual users will reduce their expenditure on IT and improve the production efficiency. Thanks to Cloud Computing, all social arenas will get IT service in lower cost and more conveniently. As the hottest IT technology, Cloud Computing is and will be always changing the whole landscape of IT industry. Predictably, Cloud Computing will lead the next wave in IT industry just like the personal computes and internet several decades ago.However, no matter how Cloud Computing developed, we need to use provide all the IT service by means of data centers. Data centers especially the ones virtualized get more and more attention. At the same time, data center consumes a large amount of energy, space and cost, and the consumption is growing day by day. As the deepening of the information service, more and more enterprises tend to construct a greener and more energy saving data center, the top challenge is to improve the usage rate of all resources in the data center and save energy, corresponding, virtual machine deployment became an important topic in reseach.In the past two years, centralized integration through virtualization is becoming a powerful way to help improve the data center utilization and save energy. Afterward, we need some Cloud Computing Platform to make a large number of servers work together, deploy and open services more conveniently, find and recover system failure, operate the whole system in a automated and intelligent manner. With the development of Cloud Computing, the Cloud Computing technology is more and more mature, one of most popular one is OpenStack. However, the virtual machine deployment mechanism of OpenStack still has some deficiencies, they are as follows:first, when we use the default scheduling algorithm, the resources on the servers used has not been fully used, some of the resources can not be used any more. Second, the default algorithm does not find the most suitable virtual machines for each server used, thus we need a larger number of servers to accommadate the virtual machines, which increase the power comsumption and the operational cost.In this passage, we proposed a new solution-virtual machine deployment strategy based on ant colony algorithm. We use this strategy to process the virtual machine request and our aim is to reduce the number of server used and cut the cost, which took advantage of the distributed coordination and positive feedback mechanism in which ants searching for food. We set up an realistic experimental environment and verified the availability and relialility of the platform which uses the strategy that we designed. At last, we did a lot of experiments to compare the different algorithms. The result shows that our strategy improves the utilization rate of the resources and reduces the number of servers used.