Bassirou Gueye ; Olivier Flauzac ; Cyril Rabat ; Ibrahima Niang - Self-adaptive structuring for P2P-based large-scale Grid environment

arima:2574 - Revue Africaine de Recherche en Informatique et Mathématiques Appliquées, December 13, 2016, Volume 25 - 2016 - Special issue CNRIA 2015 - https://doi.org/10.46298/arima.2574
Self-adaptive structuring for P2P-based large-scale Grid environmentArticle

Authors: Bassirou Gueye 1,2; Olivier Flauzac ORCID1; Cyril Rabat ORCID1; Ibrahima Niang 2

In this paper, we propose an extension and experimental evaluation of our self-adaptive structuring solution in an large-scale P2P Grid environment. The proposed specification, enables both services deployment, location and invocation of while respecting the P2P networks paradigm. Moreover, the specification is generic i.e. not linked to a particular P2P architecture. The increasing size of resources and users in large-scale distributed systems has lead to a scalability problem. To ensure the scalability, we propose to organize the P2P grid nodes in virtual communities. A particular node called ISP (Information System Proxy) acts as service directory within each cluster. On the other hand, resource discovery is one of the essential challenges in large-scale Grid environment. In this sense, we propose to build a spanning tree which will be constituted by the set of formed ISPs in order to allow an efficient service lookup in the system. An experimental validation, through simulation, shows that our approach ensures a high scalability in terms of clusters distribution and communication cost.


Volume: Volume 25 - 2016 - Special issue CNRIA 2015
Published on: December 13, 2016
Accepted on: December 8, 2016
Submitted on: December 3, 2016
Keywords: Spanning Tree,Oversim,Distributed Algorithms,Clustering,P2P Systems,Grid services,Grilles de Services,Système P2P,Algorithmes distribués,Arbre couvrant,Structuration,[INFO] Computer Science [cs]

Consultation statistics

This page has been seen 434 times.
This article's PDF has been downloaded 729 times.