Design And Implementation Of Financial Cloud Platform Host Machine Selective Algorithm That Based On Personalized Demand

In recent years, the concept of cloud computing is getting hot. Cloud is a kind of brand-new service mode, through the network and virtualization technology, the cloud platform integrates basic hardware, which allows users get away from the management of IT infrastructure and platform, and puts more energy into the business logic, it has improved the usage rate of the computing resources at the same time, it changed the service way of the traditional IT, letting both the use and maintenance of IT service go across the time and space limits.Every change in the field of IT is bound to spread to the financial sector, and the cloud computing has brought new development opportunities and new requirements to the financial sector. All the financial institutions successively upgrade and reform the big data center by using the platform of the cloud computing, so as to develop new business fields and optimize management. This has put forward higher requirements for cloud computing, which not only requires to ensure the safety of financial businesses and data, it also requires the technical force to have higher reliability. A good cloud infrastructure should provide full safety and reliability for the financial business.Cloud computing service is built on the virtual resources that are deployed on the host clusters. The traditional virtual machine deployment usually singly set the current CPU status of the host as the target host, in which the status of the host machine's other resources is not taken into account as deployment parameters, while ignoring the characteristics of the resources need to be consumed by the virtual machine. Because the business objectives need to be completed by different users is not the same, so the resource requests put forward by them also have different characteristics. Simply relying on the index of the host CPU as the basis fordeployment may cause the result that the host can not meet the user's IT resources request. With the increase of the users of the cloud platform, it may affect the load balancing of the entire cloud platform, and thus resulting in the unreasonable and ineffective use of the IT resources. Therefore, the selection of the host and the deployment of the virtual resources play a crucial role for cloud computing to provide quality services.In the paper, it proposed the host selection model which is based on personalized needs, the model quantifies the user's resource requests, then it compared the quantitative results and the performance of the host in the cloud, it deployed the virtual resources to the host by choosing the the most suitable one for the user. At the same time, it conducted an in-depth study on the load forecasting mechanism and loa balancing mechanism which are involved in the model, and it put forward the suitable load forecasting and balancing mechanism to the model.After the experimental verification, compared with the traditional host selection method, under the precondition of fully meeting the needs of individual users, at aspects of reducing the frequency of virtual machine migration, balancing the load of the cloud platform, and ensuring the stability of the system, we can see significant improvement.