Research On Fault-tolerance Mechanism For Cloud Computing Based On Virtualization Technology

With the development of cloud computing, cloud computing is playing an increasingly important role in our daily life and work. However, cloud computing service is facing the challenges of reliability and stability, and based on virtualization technology this article presents a service fault-tolerance scheme for cloud computing.In the cloud computing environment, as each service are provided by multiple virtual machines together, inappropriate deployment of virtual machines will seriously affect the performance and reliability of the services; and cloud computing resource is used by the tenants, so there are a number of new tenants to join in the cloud computing, and at the same time many other tenants due out, thus there is in significant different load on various hosts which can reduce the overall performance of the datacenter. Load balancing can be got with the migration of virtual machines, which can improve the performance of the datacenter. However, the current load balancing scheme which is based on virtual machine migration does not take into account the influence of the migration of virtual machines on service reliability. Based on the characteristics of cloud computing, based on the virtualization technologies, this paper proposes a service guaranteeing fault-tolerant solution, including the service fault-tolerance scheme based on deployment of virtual machines and service fault-tolerance scheme based on virtual machine migration. With the service fault-tolerance scheme based on deployment of virtual machines, dual-descending strategy is used to deploy the virtual machines onto the hosts, which can not only meet the performance and reliability requirements of the services, but also can save resources; with the service fault-tolerance scheme based on virtual machine migration, the impact on the performance and reliability of the services of the migration of virtual machine is evaluated to ensure that the service performance and reliability.In this paper, the effectiveness of the fault-tolerance scheme is proven with theoretical analysis, and the stability of the scheme is confirmed through experimental simulation. Simulation results show that the scheme can use less resources to ensure the performance and reliability of the services and the fault-tolerance requirements of all services are guaranteed while load balancing, and for various scale of datacenter, the scheme has been tested with a good reliability capabilities.