Machine Learning Approach for Cloud Data Analytics in IoT

9Performance Metrics for Comparison of Heuristics Task Scheduling Algorithms in Cloud Computing Platform

Nidhi Rajak^* and Ranjit Rajak

Department of Computer Science and Applications, Dr. Harisingh Gour Central University Sagar, M.P.

Abstract

Cloud computing is a recent demanding technology and infrastructure paradigm which is using in every field of science and technology. This technology is based on the Internet and apply one principle “pay and usage of computing resources”. Scheduling of the tasks is burning area of research in cloud, and it can be defined as the process of mapping of tasks onto the virtual machines and it should give overall minimum scheduling length that is minimize the overall execution time on cloud servers. Here, a very specific graph which is used and is called as Directed Acyclic Graph (DAG) represents of an application in the scheduling problem. A DAG is stated as the collection tasks and communication links with their value. This chapter studies four well-known heuristics of task scheduling algorithms such as HEFT, CPOP, ALAP, and PETS and finds their comparison studies based on performance metrics such as scheduling length, speedup, efficiency, resource utilization, and cost.

Keywords: DAG, upward rank, downward rank, cloud computing, speedup

9.1 Introduction

Technology is dynamic and it is changing present requirements which are prospective to future demands. Computing capability is growing at a rapid speed in every field and should be solved ...