Practical Aspects of Large-Scale Distributed Computing

Abstract

In the exciting and rapidly moving world of parallel and distributed computing, advances in these multi-faceted fields need to be evaluated early and disseminated quickly. The International Symposium on Parallel and Distributed Computing (ISPDC), as well as Scalable Computing—Practice and Experience (SCPE), attempts to provide interactive, professional, and friendly forums for these purposes. Following the ISPDC meetings in Lille in 2005, Cork in 2004, Slovenia in 2003, Iasi in 2002, the 5th ISPDC meeting was held in Timisoara, Romania, from July 6th to 9th 2006. Of the 84 submissions received, 45 were chosen for delivery and inclusion in the IEEE Computer Press post-proceedings. These papers were limited to eight pages, thus the authors of the best papers were invited to publish extended versions of their contributions in SCPE. Nine cutting edge papers were selected for publication in this issue.

The first paper, by Emmanuel Jeanvoine, Louis Rilling, Christine Morin, and Daniel Leprince,  presents Vigne, a Grid operating system that deals with the huge number of nodes and their dynamic behavior by using peer-to-peer overlays.

In the second paper, Ivan Frain, Abdelaziz M'Zoughi, and Jean-Paul Bahsoun introduce a dynamic quorum protocol, called the elementary permutation protocol, that permits the dynamic reconfiguration of a tree-structured coterie in function of the load of the machines that possess data replicas.

The third paper, by Ricolindo Carino and Ioana Banicescu, describes a dynamic load balancing tool intended to simplify the conversion of sequential programs containing computationally intensive one- or two-dimensional loops with independent iterates into parallel programs that execute with high efficiency on general-purpose clusters.

In the fourth paper, Miguel Bernabeu, Mariam Taroncher, Victor Garcia, and Ana Vidal describe a parallel implementation of a Lanczos-based method to solve generalized eigenvalue problems related to the modal computation of arbitrarily shaped waveguides.

Ouissem Ben Fredj and Eric Renault presents in the fifth paper of this issue the design of InfiniWrite, an implementation of the lightweight communication interface RWAPI over a Grid infrastructure called GRWA.

The sixth paper, by Keith Rochford, Brian Coghlan and John Walsh, describes the design and implementation of an agent-based architecture capable of detecting and aggregating status information using low-level sensors, functionality tests and existing information systems.

In the seventh paper, Xinfa Wei and  Kaoru Sezaki propose a new multidimensional indexing structure for P2P systems called Distributed Hilbert R-trees that enables fault-tolerant, scalable, and multidimensional range query to be executed similarly as in overlapping regions tree in P2P systems.

Maria Lopez, Elisa Heymann and Miquel Senar analyze, in the eight paper of this issue, several heuristics used as static and dynamic scheduling strategies in a Grid environment and classify them according different practical criteria.

Finally, the ninth paper, by Rocco Aversa, Beniamino Di Martino, Salvatore Venticinque and Nicola Mazzoca,  describes the functional model, the architecture design and the prototypal implementation of a platform called MAgDA, for services adaptation and delivery using agent technologies.

Note that the above mentioned papers represents different topics covered by ISPDC and SCPE that are intensively studied in the last years and we hope that the reader of this issue will find the papers interesting for his or her future research activities.

Dana Petcu