Cloud computing is a novel technology leads several new challenges to all organizations. It has been adopted by worldwide scientifically and commercially. For optimal use of cloud's potential power, effective and efficient algorithms are required, which will select best resources from available cloud resources for different applications. Allocation of user requests (tasks) to the cloud resources can optimize various performance parameters like makespan, throughput, energy consumption, etc. The task allocation or mapping problem is a well-known NP-Complete problem. Novel load balancing algorithmic approach are presented to organizing the virtualized resources of the cloud data center efficiently. The load assigned to a VM scaling up and down according to the resource available with the VM. The application of greedy algorithms are presented that minimizes the average waiting of the system, and makespan, which results in optimal use of the resource utilization and overall energy consumption. A simulation study has been presented with CloudSim to compare the performance of Task-Based allocation, First Come First Serve, Split Based Allocation and random algorithm in the cloud.
