An Architectural Framework for Scheduling and executing large scientific workflows in cloud environment

Abstract

Execution of scientific workflows in a Cloud environment is an evolving paradigm for modeling and formalize scientific applications to speed up the scientific discovering [1] process. These days’ cloud computing is getting more popular in the field of data science with which applications which involve large amount of scientific data in the pattern of scientific workflows are scheduled and executed in cloud resources. Scientists are working on large datas and trying various methods to deal with analyzing the large volume of datas. The popular examples for scientific applications are cybershake, montage, Epigenomics etc. Scheduling the execution of such scientific workflows is a major issue in the area of data science because it involves various essential supporting tasks such as management of data and task dependencies, task scheduling and execution, provenance tracking etc., to scientific applications. So, it is very important to design a scientific workflow management system for the cloud environment to overwhelm the large set of scientific data and the complexity involved in analyzing the scientific data. In this paper we proposed a referral architecture for managing the execution of scientific workflows into cloud platform which can be used as a reference model for developing application which schedules the workflow on the cloud.