Research And Design Of Resource Scheduling In Cloud Computing Based On Genetic Algorithm

Cloud computing is a more popular and advanced computing mode nowadays.Through the cloud computing,various resources such as the server can be constructed and integrated well.It also allows the users to use mass computing and storage resources conveniently.Resonable allocation of resources and QoS guarantee are the need to integrate the various resources and schedule scientifically and efficiently.They are also the need to adapt to users' different requirements and improve the users' experiences.A virtual machine scheduling method which based on genetic algorithm is introduced.It aims not only to achieve the load balancing of the server but also to minimize migration overhead.The cloud computing scheduling algorithm is improved to avoid its disadvantage of paying no attention to the changes and historical data of the virtual machines and causing load imbalance.The load balancing calculation of the serve is made better,which can analyse both the current situation of the load and the historical data and represent the load of the server mathematically.Virtualization deployment of the physical node is also done.The relationship of the virtual machine-physical machine is modeled.Not only the load of the current server is estimated,the historical load data is also analysed.A model of virtualization load constraint and mapping scheme was brought up.A virtual model has been set up and the system architecture of cloud computing has been built.The dynamic on-line scheduling algorithm for load balancing in cloud computing environment is studied and improved here.When the virtual machine resources are allocated to each physical node,the impact of resource allocation to the system will be calculated in advance by the genetic algorithm.Then it will optimize the allocation strategy to minimize the system load.The process of pre calculation and distribution is accomplished by collecting the information of historical load and current load,using genetic algorithm.According to historical load data and the current load on the system state,through the genetic algorithm,the algorithm searches an optimal allocation scheme which satisfies the load change of constraints and reduce s or avoids dynamic migration.In the efficiency of the algorithm,the load change rate and the average load distance are introduced.The former is used to describe the load change of the virtual machine,while the latter measure the global effect of load balancing.Finally,experiments are done to verify the algorithm has better effect of load balancing under the small scale nodes in a LAN environment compared to the smallest load first algorithm and round robin algorithm both in the physical constant and varying load.At the same time,it is also verified that it can still keep the minimum migration rate of the three as the number of the virtual machine increasing.The results of the experiments show that the proposed algorithm has good global convergence and resource utilization rate.It can also solve the load imbalance and large dynamic migration overhead in the cloud computing resource scheduling well.Therefore,the algorithm is feasible and effective.