Optimized Hybrid Metaheuristic Model for MapReduce Task Scheduling Applications - A Novel Framework

It takes a lot of time and work to process huge file sizes on a single processor. To meet real-time constraints, tasks must be divided into smaller subtasks and executed parallelly on multiple processors in order to be processed efficiently. MapReduce is a programming model that allows for the effective and efficient processing of tasks involving large amounts of data. It divides the large task into smaller subtasks and distributes these subtasks across a large number of VMs for execution. Scheduling these tasks on a set of available virtual machines increases execution time, consumes energy, and raises security concerns. The scheduling of large tasks under the constraints of available resources can be made more effective and efficient by optimizing these parameters of interest. A framework of the optimized hybrid metaheuristic model for MapReduce task scheduling process is presented here. It combines two well-known task scheduling algorithms, Moth Flame optimization (MFO) and Tunicate Swarm optimization (TSO). In this case, the advantages of both algorithms are combined to present an improved framework for scheduling multiple tasks on a limited number of VMs. It improves the scheduling process by optimizing important factors like make-span and cost of execution. When compared to well-known existing models, the proposed framework can optimize the important parameters of MapReduce task scheduling