The Research On Resource Scheduling With Multi-QoS Optimization In Hybrid Cloud

Cloud computing is a new type of business computing model and service model that aggregates computing,storage,content delivery,network and applications from user terminal to the cloud.Users can acquire cloud resources in a pay-as-you-go manner.On the current mode of cloud operation,cloud computing is further categorized as private cloud,public cloud and hybrid cloud.Private cloud is applied for the enterprise or organization,which has advantages such as high quality of service,security and reliability.Public cloud is a platform with a large number of computer resources as a third-party providers and it's full of large-scale and elastic scalability.With the rapid development of the big data and cloud computing,the emerging hybrid cloud has a wide range of practical value.Hybrid cloud combines the characteristics of private cloud with public cloud better.In addition,it breaks the limitation of private hardware resources.So hybrid cloud would be much more flexible and efficient to schedule large-scale peak work.However,hybrid cloud environment includes kinds of heterogeneous computer clusters.Different sizes of computer clusters there are significant differences and its user order of magnitude is very large.Hybrid cloud users are also quite different and the tasks submitted are of great diversity.So the resource scheduling in hybrid cloud which meets multi-objective requirements of users becomes a very complicated large-scale optimization problem.Also it is the focus of this paper that how to efficiently map the tasks to appropriate computing nodes based on multiple user demands.In order to solve the problem of resource scheduling in hybrid cloud environment,the paper abstracts it as a multi-objective optimization scheduling problem on large-scale heterogeneous resources.We first propose a twice clustering method to cluster the hybrid cloud resources for the purpose of realizing resources preprocessing.On the basis,we develop multi-objective resource scheduling model to achieve multi-objective task scheduling after constructing the association model of task and resource.Finally,a hybrid cloud scenario is built on CloudSim simulation platform that simulates the multi-objective requirements of the users and realizes the multi-objective optimization resource scheduling algorithm proposed in this paper.The superiority of the algorithm is illustrated by comparing experiments.This paper aims to provide a feasible solution for resource scheduling in hybrid cloud environment.