Heuristic and optimization based Scheduling algorithms for task Scheduling and load balancing in Cloud datacenters

Abstract

Task scheduling is an essential procedure in cloud computing, which is done to exploit the potential of the cloud servers. It is a mechanism which allocates the best suitable Virtual Machines (VMs) to the tasks that are submitted by the clients. Due to the heterogeneous nature of the tasks and the resources in the cloud datacenters, scheduling problem in distributed and heterogeneous systems such as cloud computing is said to be an NP-complete problem and hence remains to be a challenging issue. The main objective of the scheduling problem is to minimize the schedule length or makespan so that cloud resources can be effectively utilized. Minimizing makespan will in turn minimize the execution time, response time, and waiting time of the tasks. Utilizing heuristic-based scheduling methods will provide feasible solutions to the scheduling problem in less time. newlineThe major application of scheduling is to perform load balancing in the cloud servers. Load balancing is a process of distributing the incoming workload across the available resources. From the resource providersand#8223; aspect, this increases server utilization whereas from the clientsand#8223; aspect it reduces the response time and execution time. Obtaining a near-optimal schedule could result in a well distributed workload and also improve the scheduling metrics. Therefore, optimization-based scheduling methods have been proposed to perform load balancing in the datacenters. The present research is focused in providing heuristic-based solutions to the scheduling problem and optimization-based solutions to perform load balancing newline