Horizontal Dynamic Resource Scaling by Measuring the Impacts of Scheduling Interval in Cloud Computing

ABSTRACT

Cloud servers or Host machines are having a sufficient number of computing resources, and these resources will be shared among its client applications as per the requirement. The traffic on the cloud servers or Host machines is unpredictable, and the demand may increase or decrease with a great variations many times. Even there are lots of reasons of such variations in traffic, here we are not going to discuss such reasons. During the COVID-19 pandemic, most of the meetings, teaching–learning, documents exchanging, etc. are performed using online applications. It is noticed that the both data and processing load are increased in multiple times during the last six months on the social networking sites, and clients faced the problem of low connectivity and slow processing of applications. These cases show that there is very great requirement of load balancing of computing resources on cloud servers to maintain the smooth processing of tasks. In this paper, our main focus is to measure the impacts of scheduling interval on dynamic resource allocation and to calculate the demand of resources dynamically on the basis of dynamic creation or deletion of virtual machines to serve the cloudlet load efficiently. Dynamic resource scaling can be done either horizontally or vertically. In horizontal dynamic resource scaling, the Host machines are equipped with the more number of resources rather than increasing the number of Host machines. In cloud computing, the service provider’s intention is to provide a cost-effective and efficient service support to its clients. The resources and services in cloud computing are on the pay-per-use basis, and hence, the horizontal dynamic resource scaling provides a good cost-effective solution for the resource requirement. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.