Research On Resource Allocation Optimization Model And Algorithm Of Cloud Migration For Enterprise Elastic Application

The traditional way of business management is to choose a special server to respond to tThe traditional business management way of an enterprise is to choose a specified server to respond to the service requests,or to set up its own computer room,or to rent an independent server.In recent years,with the rapid growth of enterprise business data,cloud computing technologies emerge.Cloud computing has attracted the attention and application of enterprises because of its advantages such as simple and convenient use,reducing a large number of software and hardware costs of enterprises,and saving their own operation and maintenance costs.Therefore,more and more enterprises migrate or directly deploy applications to the cloud.Under the premise of certain cloud data center resources,how to complete the migration of enterprise applications into the cloud and realize the effective deployment in the cloud is a very important research direction in the field of enterprise applications.This thesis mainly studied the resource allocation of elastic enterprise application migration into the cloud,including the bandwidth resource allocation for elastic application migration and the cloud resource allocation for elastic application deployment into the cloud.This thesis established two resource allocation models,analyzed the optimal solution of these models by using nonlinear programming theory,and designed the gradient-based resource allocation algorithms.The main contents of this thesis are as follows:First of all,this thesis considered the completion time of migration from the perspective of users,and established a bandwidth resource allocation model for elastic application migration into the cloud,which aims at minimizing the completion time.In order to achieve the optimal bandwidth allocation,a gradient-based bandwidth resource allocation algorithm was proposed and the specific implementation steps of this algorithm were discussed.Secondly,this thesis considered the migration utility of applications(i.e.,user satisfaction)from the perspective of user experience,and analyzed the operation cost of the cloud center to allocate cloud resources for the applications from the perspective of cloud resource providers.Then this thesis proposed the cloud resource allocation modelfor applications deployment into the cloud,which aims to maximize the migration utility of applications and minimize the operation cost of cloud resource providers.In order to realize the optimal resource allocation,a gradient-based cloud resource allocation algorithm was presented and the specific implementation steps of this algorithm were also discussed.Finally,this thesis introduced the implementation framework for migration and deployment of elastic enterprise applications into the cloud,and formulated three different scenarios of applications migration.Then,numerical simulation experiments for bandwidth resource allocation and cloud resource allocation were conducted respectively for migration and deployment of elastic enterprise applications into the cloud,which verify the convergence and effectiveness of the proposed resource allocation schemes.