Study Of A Flexible Load Balancing Mechanism Of Cluster On Cloud

With the development of modern computer hardware technology, making large-scale computing become possible, on this basis, in recent years raised many new computing model, cloud computing is one of them. It allows users to cheap cost in exchange for the need of IT infrastructure services, such as computing resources, storage resources. Cloud computing convenient, fast, cheap rented had a significant impact on the field of IT. Elastic cloud computing using virtualization technology to provide services centrally manage a group of distributed hardware resources, which need to solve an important problem is load balancing problems, the pros and cons of the load balancing mechanism with the quality of service provided by the cloud computing platform closely linked.Traditional cluster environment, the task processing unit is a physical host, if the host Additions excessive, not only the huge investment, and when the load is too low, it will cause the waste of computing resources; if the host is too small, the load exceeds the total load capacity, QoS caused a great deal of influence, traditional load balancing is a rigid load balancing. The resources of the virtual machine in the cloud environment is a flexible computing resources can be based on the actual load changes dynamically add or reduce virtual machine.In response to these issues, this paper presents different load capacity demand for personalized computing resource allocation, combined with load forecast method to create configure the virtual machine has a certain number of different types of resource pools elastic load balancing mechanism, when providing services request a virtual machine, the virtual machine load is too high to join the service cluster, reducing the excessive load; provide service virtual machine load is too low, you select a virtual machine offline. This paper also attempts to draw on the idea of "green", reduce the power consumption of the Cloud Computing Center, and to improve the utilization of computing resources, thereby reducing greenhouse gas emissions to the natural environment.