On Solving the Multi-Objective Software Package Upgradability Problem

On Solving the Multi-Objective Software Package Upgradability Problem

Noureddine Aribi (Lab. LITIO, University of Oran 1, Oran, Algeria) and Yahia Lebbah (Lab. LITIO, University of Oran 1, Oran, Algeria)
DOI: 10.4018/978-1-7998-9158-1.ch007
OnDemand PDF Download:
Available
$33.75
List Price: $37.50
10% Discount:-$3.75
TOTAL SAVINGS: $3.75

Abstract

Free and open source software (FOSS) distributions are increasingly based on the abstraction of packages to manage and accommodate new features before and after the deployment stage. However, due to inter-package dependencies, package upgrade entails challenging shortcomings of deployment and management of complex software systems, inhibiting their ability to cope with frequent upgrade failures. Moreover, the upgrade process may be achieved according to some criteria (maximize the stability, minimize outdated packages, etc.). This problem is actually a multi-objective optimization problem. Throughout the article, the authors propose a Leximax approach based on mixed integer linear programming (MILP) to tackle the upgradability problem, while ensuring efficiency and fairness requirements between the objective functions. Experiments performed on real-world instances, from the MANCOOSI project, show that the authors' approach efficiently finds solutions of consistently high quality.
Chapter Preview
Top

2. Preliminaries

Before discussing our approach, we provide some necessary background.

Complete Chapter List

Search this Book:
Reset