Optimal Load Distribution by Queue Partitioning Technique in Cloud Environment

The ambitious nature of the present digital era demands the IT industries to increase their return on investment and achieve better results. With this in mind, technologies have been developed and put into use. One such emerging technology that is making a big impact is cloud computing. Being on cloud is a benefactor to businesses and IT sector to save resources, time, investment and maintenance cost. This has led to adoption of cloud services by most of the organizations and the demand for cloud computing technology has observed steep increase with time. As a result, due to greater usage of cloud services there has been an increase in workload leading to an unequal grouping of work load among servers resulting in server being overloaded, affecting its performance. The challenge now is to assure Quality of Service (QoS) to customers by equal distribution of workload among servers. This issue is addressed in this paper by providing a solution to balance the load among the servers on the cloud and hence amplify the Quality of Service. This paper puts forward an algorithm - “Queue Partitioning Technique” where the incoming tasks queue are partitioned and tasks are distributed to virtual machines based on concept of quick sort technique. The suggested algorithm uses waiting time and load distribution as a QoS parameter and compared with two existing techniques. Proposed algorithm is implemented and evaluated using CloudSim simulator.