Our apologies if you receive multiple copies of this announcement.
********************************************************************** * ASSESS * * Workshop on Assessing Models of Networks and * * Distributed Computing Platforms * * May 19-22, 2008 * * Lyon, France * * http://www-id.imag.fr/Laboratoire/Membres/Legrand_Arnaud/assess08/ * * * * Workshop held in conjunction CCGrid 2008 * * http://www.ens-lyon.fr/LIP/RESO/ccgrid2008/ * **********************************************************************
Paper Submission Deadline: November 19, 2007 --------------------------------------------
Description
The Workshop on "Assessing Models of Networks and Distributed Computing Platforms" (ASSESS) is a one-day event held in conjunction with the 8th IEEE International Symposium on Cluster Computing and the Grid ([7]CCGrid 2008), which will be held in Lyon, France, on May 19th, 2008.
Motivation
Models are the organization of thoughts and facts through more or less simple laws. These models can then be used to think about reality or to make predictions. Therefore, scientists strive for objective models and knowledge based on objective observations.
With the first computers, or even before, computer scientists have designed computing models, from the architectural point of view as the classical Von Neuman model, to complexity models like the Turing Machine. Then, with the parallel machines there appeared several computing models from the first ones, such as PRAM, and ranging from more architectural ones like LogP, to more conceptual ones such as CGM/BSP. These models were adapted with some changes to the cluster computing. There is still the need though for specific models for large-scale distributed platforms like grids or peer-to-peer infrastructures where heterogeneity and changes are usual. Wide area networks are crucial components of such platforms and thus raise the same issues. Clearly, a "good" model should be: * Realistic; * Tractable; * Instantiable.
Evaluating the quality of a model is tricky. In particular because results of computer measurements are generally brittle and sensitive to initial conditions. Moreover due to the rapid evolution of technology and systems, measurements and analysis often quickly become obsolete. As a consequence, computer measurements do not lead to a universal and enduring truth.
This difficulty is not a curse but computer scientists should thus be even more careful and precise than any other scientists. In particular they should regularly check the validity of their models and precisely detail the circumstances under which the data they use is collected. A particular emphasis should be put on providing enough information to enable others to reproduces results. This good practice is unfortunately not very common. Most people often use models (either from their own or from others) without ever questioning their relevance. Many projects and platforms such as PlanetLab, Grid5000, DAS3 enable people to perform large scale experiments and check the validity of their hypothesis. However resorting to simulation is very common as driving real life large-scale experiments is very tedious. This may be a reasonable option but very few simulation users and developers question the validity of the models underlying their simulators.
Coming up with validated models and simulators is thus very difficult, especially when dealing with large-scale distributed infrastructures like grids or peer-to-peer infrastructures.
Goal
This workshop aims to provide advances on models for distributed computing (e.g. grids, clusters, network, peer-to-peer infrastructure, etc.), gathering mainly experimental work. The workshop welcomes analysis, experience reports, simulations, emulation, performance evaluation techniques; Proposals for model extensions, as well as proposals for innovative models accompanied with experiments are welcome. A particular attention will be paid to methodology and reproducibility of results.
This workshop also aims to be an opportunity to meet people and have live discussions on this hot topic. The topics of interest include, but are not limited to, experiments, measurements, performance analysis, evaluation, validation and experiences on: * Communication Models * Network Models (contention, topology, traffic...) * Storage Models * Programming Models * Benchmark, Profiling and Forecast * Workload * Scalable Models * Simulation, Emulation * Global Computing * Peer-to-Peer * Grid Computing * Cluster Computing
All workshop papers are a maximum of 6 pages in length (in IEEE format). The publication of proceedings will be by the IEEE alongside the CCgrid conference proceedings.
At least one author of each accepted submission must attend the workshop and all workshop participants must pay the CCGrid 2008 workshop registration fee, as well as the conference fee.
Furthermore, authors of selected high quality papers will be invited to submit revised and extended versions for publication in a special issue in a higly reputation journal (we consider contacting Parallel Computing).
PC chairs:
* Alfredo Goldman (Universidade de Sao Paulo - Brazil) * Arnaud Legrand (CNRS, LIG, Grenoble, France) * Martin Quinson (INRIA/LORIA, Vandoeuvre les Nancy, France) * Olivier Richard (LIG, Grenoble, France)
Program Committee:
* Alberto Montresor (University of Trento, Italy) * Cristina Boeres (Universidade Federal Fluminense - Brazil) * Dror Feitelson (Hebrew University, Jerusalem, Israel) * Francisco Brasileiro (Universidade de Campina Grande - Brazil) * Franck Cappello (INRIA Futurs and LRI, France) * Henri Casanova (University of Hawaii, Manoa) * Jack Dongarra (University of Tennessee, Knoxville, USA) * Jay Lepreau (University of Utah, USA) * Marco Danelutto (Universita de Pisa, Italy) * Marinho Barcellos (Unisinos - Brazil) * Mitsuhisa Sato (University of Tsukuba, Japan) * Mohamed Jemni (Universite de Tunis, Tunisia) * Olivier Aumage (INRIA, Bordeaux, France) * Olivier Dalle (INRIA, Sophia Antipolis, France) * Paulo Goncalves (INRIA, ENS Lyon, France) * Thilo Kielmann (Vrije Universiteit, Netherlands) * Wolfgang Nagel (TU Dresden, Germany)
Important dates
* Paper Submission Deadline: November 19, 2007 * Notification of Acceptance: January 15, 2008 * Camera-ready Manuscripts due: January 30, 2007 * Conference Dates: To determine in May 19-22, 2008
Regards, Marinho.
Prof. Marinho P. Barcellos (PIPCA/UNISINOS) marinho@acm.org, http://p2p-sec.org
_______________________________________________ Tccc mailing list Tccc@lists.cs.columbia.edu https://lists.cs.columbia.edu/cucslists/listinfo/tccc