An Efficient Hybrid Load Balancing Algorithm for Heterogeneous Data Centers in Cloud Computing
Nageswara Prasadhu N · International Journal of Advanced Trends in Computer Science and Engineering · 2020
Cloud computing has recently grown into a major global trend of computing.Using the Internet and Wide Area Network (WAN) to make services remotely is a modern design.This is a new solution and technique to achieve high availability, versatility, cost savings and demand-scalability.However, cloud computing faces many problems, such as wasteful resource use, which has a major effect on the performance of cloud computers.These issues have to be managed from time to time to avoid the utilization factors of the various attributes that are used while implementing the process.Because of the enormous amount of knowledge these issues emerged and are unaddressed for so many years even though they were just adjusted in between to carry out the normal activities.Therefore, one of the most critical issues in this area in improving cloud computing performance is the need for robust and efficient load balancing algorithms for cloud computing.Many researchers have proposed different load balancing and job scheduling algorithms in cloud computing, but system efficiency is still very unstable and load still unbalance.This has subsequently delayed the process of executing the algorithms within the required timelines.Hence, in this research, we propose a load balancing algorithm to improve performance and efficiency in a heterogeneous cloud computing environment.We propose a hybrid algorithm that utilizes both random and greedy algorithms.The algorithm takes into account the current resource data and the CPU capacity factor to attain the objectives.The hybrid algorithm was tested using Cloud Analyst simulator, and compared with other algorithms.This comparison has been carried out both in a subjective way and also objective way to establish the proposed method.The experimental investigations showed improvements in average response time and processing time by taking current resource information and the CPU capacity factor compared to other algorithms into consideration.