A Novel Heterogenous Dominant Sequence Clustering for Task Scheduling and Optimal Load Balancing Using JSO in Cloud

Abstract

Cloud computing (CC) is rapidly increasing and being used more and more in information technology (IT) contexts. Among the hottest issues in the world of CC is task scheduling. In cloud datacentres, load balancing is accomplished using a variety of scheduling techniques, although lengthening the parallel time. By taking task duration and VM capacity into account, this study proposes a cluster-based task scheduling paradigm. The makespan and execution duration should be reduced, the suggested system engages in dynamic load balancing. In this study, we suggest a unique method called HDDSC-JSO, which employs the Jellyfish Swarm Optimization (JSO) algorithm for load balancing and heterogeneous Density based dominant sequence clustering (HDDSC) for job scheduling. Using the HDSC technique, It presents a graph of one or more groups representing user tasks, the tasks of users are first clustered. The Adapted Diverse Expeditious Termination Time (ADETT) method is used to score each work after task clustering. whenever the job with the greatest priority is scheduled first. Next, using a JSO algorithm, load balancing is carried out, distributing the load according to the weight and capacity of the server as well as the connection of the client to the server. Task allocation chooses a heavily weighted or unconnected server, lengthening the response time. Finally, measures including response time, makespan, resource consumption, and service dependability were used to assess the suggested architecture.