Deployment Planning of Component-Based Distributed Applications Using Mathematical Programming

Abstract

Component deployment directly affects the performances of component-based distributed applications, such as response time, throughput, reliability, and security. It also affects resources utility of runtime environment. In order to reduce the cost of resources used by applications and improve the resources utility, this paper defined component graph to represent component-based distributed applications, and used tree network topology to describe the runtime environment of distributed applications. We defined the resource cost objective function and formulated component deployment optimization problem as mathematical programming problem. Example analysis demonstrated our approach was effective.

A multi-scale modeling approach for software architecture deployment

“…Other research activities have proposed architecturebased approaches using ADL (Architecture Description Language) [7,9] and graphs [5,15,14]. Hoareau et al [7] present a support for deploying and executing an application built with hierarchical components.…”

“…They demonstrate that this deployment problem corresponds to the multiway cut problem in graph theory. Also, the work of Zhang et al [15] defines a component graph to represent component-based distributed applications and a tree network topology to describe the runtime environment. It defines the resource cost objective function and formulates component deployment optimization problem as mathematical programming problem.…”

“…The generated deployment plan is dynamically executed by a mobile agent system. We can note that the research activities [7,9,5,15,1,10] deal only with resource constraints during the construction of the deployment architecture. They do not take into account the respect of structural constraints to validate he deployment architecture.…”

A formal method for modeling deployment architectures based on bigraphs

“…Other research activities have proposed architecture-based approaches using ADL (Architecture Description Language) [Hoareau andMahéo, 2006, Malek et al, 2012] and graphs [Heydarnoori and Mavaddat, 2006, Zhang et al, 2010, Bouassida Rodriguez et al, 2008. Hoareau et al [Hoareau and Mahéo, 2006] present a support for deploying and executing an application built with hierarchical components.…”

“…They demonstrate that this deployment problem corresponds to the multiway cut problem in graph theory. Also, the work of Zhang et al [Zhang et al, 2010] defines a component graph to represent component-based distributed applications and a tree network topology to describe the runtime environment. It defines the resource cost objective function and formulates component deployment optimization problem as mathematical programming problem.…”