Multi-Objective Optimization of Deadline and Budget-Aware Workflow Scheduling in Uncertain Clouds

Maria Carla Calzarossa,Daniele Tessera,Marco L Della Vedova,Giuseppe Nebbione,Luisa Massari

doi:10.1109/access.2021.3091310

Maria Carla Calzarossa, Daniele Tessera + Show 3 more

Open Access

https://doi.org/10.1109/access.2021.3091310

Copy DOI

Abstract

Cloud technologies are being used nowadays to cope with the increased computing and storage requirements of services and applications. Nevertheless, decisions about resources to be provisioned and the corresponding scheduling plans are far from being easily made especially because of the variability and uncertainty affecting workload demands as well as technological infrastructure performance. In this paper we address these issues by formulating a multi-objective constrained optimization problem aimed at identifying the optimal scheduling plans for scientific workflows to be deployed in uncertain cloud environments. In particular, we focus on minimizing the expected workflow execution time and monetary cost under probabilistic constraints on deadline and budget. According to the proposed approach, this problem is solved offline, that is, prior to workflow execution, with the intention of allowing cloud users to choose the plan of the Pareto optimal set satisfying their requirements and preferences. The analysis of the combined effects of cloud uncertainty and probabilistic constraints has shown that the solutions of the optimization problem are strongly affected by uncertainty. Hence, to properly provision cloud resources, it is compelling to precisely quantify uncertainty and take explicitly into account its effects in the decision process.

Highlights

Cloud infrastructures are the computing environments commonly used nowadays to deploy distributed applications and services
EXPERIMENTAL RESULTS This section investigates the effects of cloud uncertainty on the solutions of the constrained optimization problem
CLOUD UNCERTAINTY To investigate the effects of cloud uncertainty, we designed several experiments varying the type of probability distributions (i.e., Uniform, Half-Normal and Weibull) describing the cloud characteristics (i.e., Virtual Machine (VM) processing capacity, data transfer rate, network bandwidth)

Summary

Introduction

Cloud infrastructures are the computing environments commonly used nowadays to deploy distributed applications and services These technologies offer many benefits, including, among the others, reduced costs, scalability and flexibility. Cloud resources can be rapidly and elastically scaled up or down as needed These features – combined with the utility-based pricing model of the resources – make cloud computing attractive for enterprises and organizations that can avoid the burden of buying, installing and maintaining their own infrastructures. In these complex scenarios, users are willing to devise the most cost effective solution able to satisfy the requirements of their workloads [1], [2]. It is up to the users to choose the provider (or providers) and decide about the quantity (e.g., number of Virtual Machines) and characteristics (e.g., processing capacity) of resources to be provisioned for deploying their workloads

Objectives

Methods

Results

Conclusion