An Approach for the Load Balancing in Cloud Based on the Dynamic Threshold

Cloud computing offer IT services to the users worldwide on the basis of pay-as-you-go model. Now a day cloud is so famous because of their attractive features such as universal accessibility, availability, cheap, pay as you go etc. Cloud is a business model, where the cloud provider, gives the resources on rent basis. Cloud provider wants to maximum profit, which can be achieved by the proper resource utilization and reduction in energy consumption. Energy consumption is depends on the number active server [1] says that a host consume 70 % (of their total energy consume when they are fully utilize) even when the host is completely idle. So energy consumption can be reduced by minimizing the number of active server. Load balancing is one of the critical issue in the cloud, because it affect the user performance. In this paper we proposed a adaptive threshold based load balancing approach which increase the resource utilization and reduce the energy consumption. This approach uses two thresholds that are upper and lower threshold for the load balancing. Load of the physical machine (PM) must be within the lower and upper threshold rang. When the load on the host is above/below the threshold, some VM will be migrated to balance the load on the host.