Dynamic Resource Allocation with Efficient Power Utilization in Cloud

In the Cloud, the most challenging task is to allocate resources for jobs. The major objective is to allocate resources to various tasks by reducing the CPU utilization of physical machines in the cloud. The various multi-agents look after the incoming jobs, schedules the jobs to various host machines, dynamically allocates jobs or tasks in other machines in order to reduce CPU power. VM migration is done through swapper agent when a host CPU utilization rises over the threshold value and if other hosts are working in less CPU power. So the agents can be used to migrate VMs from one node to another in run time, so that it can be stabilized and CPU utilization can be reduced. To increase the efficiency during allocation of virtual machine for jobs, a threshold based best fit algorithm is proposed to sort the VMs in the increasing order of CPU utilization. The power utilization is based on the number of hosts working at a certain period and the proposed strategy is used to reduce the power to improve the resource utilization in Cloud.