Desktop Grid Computing
|Editors||C. Cérin & G. Fedak|
The goal of the Desktop Grid Computing is to offer a synthesis on various approaches in Desktop Grid Computing, a decade after the pioneering work. Desktop Grids use computing, network and storage resources of idle desktop PCs distributed over multiple LANs or the Internet. Today, this type of computing platform forms one of the largest distributed computing systems, and currently provides scientists with tens of TeraFLOPS from hundreds of thousands of hosts. Several topics will be discussed in details, which may appeal to a much larger audience than the community of researcher in the field. DGC is certainly the first book that makes a deep comprehensive and up-to-date presentation about the topic. It also draws perspectives, recent trends in the domain that underline the evolution of the initial concept.
The book is intended primarily for researchers, advanced undergraduate and graduate students in distributed computing, grid computing, scientific computing, high performance computing and related fields, as well as to engineers, professionals from academia and industry. The proposed book can be used as a textbook for a graduate course, for an advanced undergraduate course, or for a summer school. It will serve as a self-contained book and detailed overview of the relevant Desktop Grid models. The general area of Desktop Grid computing and its state of-the-art methods and technologies are essential to many other computational science fields, including health, bio-informatics, medical imaging, physics, astronomy, climate prediction, mathematics, cryptography and more generally to simulation.
About the book:
The book will be devoted to survey common techniques used in numerous models, algorithms, tools developed during the last decade to implemented the concept of Desktop Grid computing. Such techniques allow to implement and to solve many important sub-problems for middleware design. Such sub-problems are: scheduling, data management, security, volunteer computing, load balancing, model of parallelism, programming models, result certification, fault tolerance. The book will cover approximately X-Y pages of methods and applications, with accompanying tables and illustrations. It will seek to balance the theory of designing Desktop Grid middleware and architecture, with practice, applications and real world deployment on large scale platforms. The book is organized in two parts. The first part is related to the birth (initial ideas, basic concepts...) of Desktop Grid Computing and the second part is related to what we call 'the maturity' of Desktop Grids that is to say to a more prospective vision of challenging problems for present and the future.
Each chapter will include the presentation of the sub-problems addressed in the chapter, discussions on the theoretical and practical issues, details about chosen implementation and experiments and finally references to further reading and notes.
The "Introduction" will give an overview of the problems to be discussed in the book, a brief history of various models in Desktop Grid Computing and examples of applications. The key organization of the two distinct parts is mainly according to a dedicated middleware, applications, experiments for the first part without forgetting the key essential notions (scheduling, data management, security, load balancing, model of parallelism, programming models, result certification, fault tolerance). The second part explores areas not covered in the first part, mainly not yet implemented topics and that constitutes research topics in themselves.
Table of Contents
Introduction (Christophe Cérin, Gilles Fedak)
Forewords (Franck Cappello)
Part I: The Birth
- Chapter 1 Volunteer Computing and BOINC David Anderson
- Chapter 2 Open, Scalable and Self-Regulated Federations of Desktop Grids with OurGrid Francisco Brasileiro and Nazareno Andrade
- Chapter 3 The XtremWebCH Volunteer Computing Platform Nabil Abdennadher, Marko Niinimaki, and Mohamed BenBelgacem
- Chapter 4 XtremWeb-HEP: Designing Desktop Grid for the EGEE Infrastructure Oleg Lodygensky, Etienne Urbah and Simon Dadoun
- Chapter 5 A Volunteer Computing Platform Experience for Neuromuscular Disease Problems Nicolas Bard, Viktors Bertis, Raphael Bolze, and Frédéric Desprez
- Chapter 6 How to Work with XtremWeb, Condor, BOINC on Top of BonjourGrid Christophe Cérin, Heithem Abbes, and Walid Saad
- Chapter 7 How to Work with PastryGrid Christophe Cérin and Heithem Abbes
- Chapter 8 Towards Clouds@home: Integration of Virtualization in Desktop Grids with OAR and BOINC Derrick Kondo and Thiago Presa
Part II: The Maturity and Beyond
- Chapter 9 Challenges in Designing Scheduling Policies in Volunteer Computing Trilce Estrada and Michela Taufer
- Chapter 10 Modeling and Optimizing Availability of Non-Dedicated Resources Artur Andrzejak and Derrick Kondo
- Chapter 11 Security and Result Certification Filipe Araujo and Patrício Domingues
- Chapter 12 Data-Intensive Computing on Desktop Grids Heshan Lin, Gilles Fedak, and Wu-chun Feng
- Chapter 13 Roles of Desktop Grids in Hybrid Distributed Computing Infrastructures Simon Delamare and Gilles Fedak
- Chapter 14 Supporting Web 2.0 Communities by Volunteer Desktop Grids Kacsuk Peter, Attila Marosi, Robert Lovas, and Jozsef Kovacs
- Chapter 15 Programming Applications for Desktop Grids Tamas Kiss and Gabor Terstyanszky
- Chaper 16 Network Awareness in Volunteer Networks Jon B. Weissman and Jinoh Kim