Optimizing Greedy Algorithm to Balance the Server Load in Cloud Simulated Environment

Abstract

The performance of cloud services depends on the scheduling algorithms that distribute the incoming network traffic among their servers to achieve effectiveness in execution of tasks. These algorithms are assigning the tasks to various computing resources, and these resources are virtual in nature. In cloud, assigning tasks to corresponding resources are NP-hard in nature. The traditional scheduling algorithms like FCFS, SJF, Round Robin etc. will not be suitable to solve NP-hard scheduling problems. Cloud scheduling considers various criteria like resource utilization, cost, makespan and throughput. This paper has implemented the cloud scheduling algorithms such as Max-Min Algorithm, Min-Min Algorithm, Enhanced Max-Min Algorithm and Greedy Algorithm to balance the server load in cloud environment and have analyzed the results of these algorithms to identify the best scheduling algorithm. Results discussed in this paper have shown that, when the numbers of tasks are more, greedy algorithm outperform other scheduling algorithms while for less number of tasks, Enhanced Max-Min algorithm performs extremely well as compared to another task scheduling algorithm.