Integrated Research in GRID Computing- P1 pps

Integrated Research in
GRID Computing
Integrated Research in
GRID Computing
CoreGRID Integration Workshop 2005
(Selected Papers)
November 28-30, Pisa, Italy
edited by
Sergei Gorlatch
University
ofMUnster
Germany
Marco Danelutto
University
of Pisa
Italy
Springer
Sergei Gorlatch Marco Danelutto
Universitat Munster Dept. Computer Science
FB Mathematik und Informatik University of Pisa
Inst. f. Informatik Largo Pontecorvo, 3
Einsteinstr. 62 56127 PISA
48149 MUNSTER ITALY
GERMANY

Library of Congress Control Number: 2006934290
INTEGRATED RESEARCH
IN
GRID COMPUTING
edited by Sergei Gorlatch and Marco Danelutto

ISBN-13:
978-0-387-47656-3
ISBN-10: 0-387-47656-8
e-ISBN-13: 978-0-387-47658-2
e-ISBN-10: 0-387-47658-X
Printed on acid-free paper.
© 2007 Springer Science+Business Media, LLC
All rights reserved. This work may not be translated or copied in whole or
in part without the written permission of the publisher (Springer
Science+Business Media, LLC, 233 Spring Street, New York, NY 10013,
USA),
except for brief excerpts in connection with reviews or scholarly
analysis. Use in connection with any form of information storage and
retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now know or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks and
similar terms, even if the are not identified as such, is not to be taken as
an expression of opinion as to whether or not they are subject to
proprietary rights.
Printed in the United States of America.
987654321
springer.com
Contents
Foreword vii
Contributing Authors xi
Data integration and query reformulation in service-based Grids 1
Carmela Comito and Domenico Talia, Anastasios Gounaris and Rizos Sakellariou
Towards a common deployment model for Grid systems 15
Massimo Coppola and Nicola Tonellotto, Marco Danelutto and Corrado Zoccolo,
Sebastien Lacour and Christian Perez and Thierry Priol

Towards Automatic Creation of Web Services for Grid Component Composition 31
Jan DUnnweber and Sergei
Gorlatc/%
Nikos Parlavantzas, Francoise Baude and Virginie
Leg rand
Adaptable Parallel Components for Grid Programming 43
Jan DUnnweber and Sergei Gorlatch, Marco Aldinucci, Sonia Campa and Marco
Danelutto
Skeleton Parallel Programming and Parallel Objects 59
Marcelo Pasin, Pierre Kuonen, Marco Danelutto and Marco Aldinucci
Towards the Automatic Mapping of ASSIST Applications for the Grid 73
Marco Aldinucci, Anne Benoit
An abstract schema modeling adaptivity management 89
Marco Aldinucci and Sonia Campa and Massimo Coppc
Corrado Zoccolo, Francoise Andre and Jeremy Buisson
Marco Aldinucci and Sonia Campa and Massimo Coppola and Marco Danelutto and
\nare
A Feedback-based Approach 103
Charis Papadakis, Paraskevi Fragopoulou, Elias Athanasopoulos, and Evangelos P
Markatos, Marios Dikaiakos, Alexandros Labrinidis
Fault-injection and Dependability Benchmarking 119
William Hoarau and Sebastien Tixeuil, Luis Silva
User Management for Virtual Organizations 135
Jiri Denemark, Ludek Maty ska, Miroslav Ruda, Michal Jankowski, Norbert Meyer,
Pawel Wolniewicz
vi INTEGRATED RESEARCH IN GRID COMPUTING
On the Integration of Passive and Active Network Monitoring in Grid Systems 147
Sergio
Andreozzi,
Augusto

Ciuffoletti,
Antonia
Ghiselli,
Demetres
Antoniades, Michalis
Polychronakis, Evangelos
P.
Markatos, Panos Trimintzios
New Grid Monitoring Infrastructures 163
Piotr Domagalski and Krzysztof
Kurowski
and Ariel Oleksiak and Jarek Nabrzyski,
Zoltdn Balaton and Gdbor Gombds and
Peter
Kacsuk
Towards Semantics-Based Resource Discovery for the Grid 175
William Groleau, Vladimir
Vlassov,
Konstantin Popov
Scheduling Workflows with Budget Constraints 189
Rizos Sakellariou and Henan Zhao, Eleni Tsiakkouri and Marios D. Dikaiakos
Integration of ISS into the VIOLA Meta-scheduHng Environment 203
Vincent
Keller,
RalfGruber, Michela Spada, Trach-Minh
Tran,
Kevin Cristiano, Pierre
Kuonen,
Philipp
Wieder,

Wolfgang
Ziegler,
Oliver
Wdldrich,
Sergio Maffioletti, Marie-
Christine
Sawtey,
Nello Nellari
Multi-criteria Grid Resource Management using Performance Prediction 215
Krzysztof
Kurowski,
Ariel
Oleksiak,
and
Jarek
Nabrzyski,
Agnieszka
Kwiecieii,
Marcin
Wojtkiewicz,
and Maciej Dyczkowski, Francesc
Guim,
Julita Corbalan, Jesus Labarta
A Proposal for a Generic Grid Scheduling Architecture 227
Nicola
Tonellotto,
Ramin
Yahyapour,
Philipp Wieder
GRID superscalar enabled P-GRADE portal 241

Robert
Lovas,
Gergely Sipos and
Peter
Kacsuk,
Raill
Sirvent,
Josep
M.
Perez
and
Rosa
M. Badia
Redesgining the SEGL PSE: A Case Study of Using Mediator Components 255
Thilo Kielmann and Gosia
Wrzesinska,
Natalia Currle-Linde and Michael Resch
Synthetic Grid Workloads with Ibis, KOALA, and GrenchMark 271
Alexandru losup and Dick HJ. Epema, Jason Maassen and Rob van Nieuwpoort
Author Index 285
Foreword
This volume is a selection of best papers presented at the CoreGRID Inte-
gration Workshop 2005 (CGIW'2005), which took place on 28-30 November
2005 in Pisa, Italy,
The workshop was organised by the Network of Excellence CoreGRID
funded by the European Commission under the sixth Framework Programme
IST-2003-2.3.2.8 starting September 1st, 2004 for a duration of four years.
CoreGRID aims at strengthening and advancing scientific and technological
excellence in the area of Grid and Peer-to-Peer technologies. To achieve this
objective, the network brings together a critical mass of well-established re-

searchers (145 permanent researchers and 171 PhD students) from forty two
institutions who have constructed an ambitious joint programme of activities.
The goal of the workshop is to promote the integration of the CoreGRID
network and of the European research community in the area of Grid and P2P
technologies, in order to overcome the current fragmentation and duplication
of efforts in this area.
The list of topics of Grid research covered at the workshop included but was
not limited to:
• knowledge & data management;
• programming models;
• system architecture;
• Grid information, resource and workflow monitoring services;
• resource management and scheduling;
• systems, tools and environments;
• trust and security issues on the Grid.
Priority at the workshop was given to work conducted in collaboration between
partners from different research institutions and
to
promising research proposals
that can foster such collaboration in the future.
The workshop was open to the participants of the CoreGRID network and
also to the parties interested in cooperating with the network and/or, possibly
joining the network in the future.
viii
INTEGRATED RESEARCH IN GRID COMPUTING
The Programme Committee who made the selection of papers included:
Sergei Gorlatch, University of Muenster, Chair
Marco Danelutto, University of Pisa
Domenico Laforenza, ISTI-CNR
Uwe Schwiegelshohn, University of Dortmund

Thierry Priol, INRIA/IRISA
Artur Andrzejak, ZIB
Vladimir Getov, University of Westminster
Ludek Matyska, Masaryk University Brno
Domenico Talia, University of Calabria
Ramin Yahyapour, University of Dortmund
Norbert Meyer, Poznan Supercomputing and Networking Center
Pierre Guisset, CETIC
Wolfgang Ziegler, Fraunhofer-Institute SCAI
Bruno Le Dantec, ERCIM
The Workshop Organising Committee included:
Marco Danelutto, University of Pisa
Martin Alt, University of Muenster
Sonia Campa, University of Pisa
Massimo Coppola, ISTI/CNR
All papers in this volume were additionally reviewed by the following external
reviewers whose help we gratefully acknowledge:
Ali Anjomshoaa
Rajkumar Buyya
Andrea Clematis
Massimo Coppola
Rubing Duan
Vincent Englebert
Eitan Frachtenberg
Dieter Kranzlmueller
Salvatore Orlando
Carles Pairot
Hans-Werner Pohl
Uwe Radetzki
Wolfgang Reisig

Michal Sajkowski
Volker Sander
Mumtaz Siddiqui
Anthony Sulistio
Hong-Linh Truong
FOREWORD ix
We gratefully acknowledge the support from the members of
the
Scientific Ad-
visory Board and Industrial Advisory Board of CoreGRID, and especially the in-
vited speakers John Easton (IBM Grid Computing
UK)
and Uwe Schwiegelshohn
(University of Dortmund). Special thanks are due to the authors of all submitted
papers, the members of the Programme Committee and the Organising Commit-
tee,
and to all reviewers, for their contribution to the success of this event. We
are grateful to the University of Pisa for hosting the Workshop and publishing
its preliminary proceedings.
Muenster and Pisa, July 2006
Sergei Gorlatch and Marco Danelutto (workshop organizers)
Thierry Priol (Scientific Coordinator of CoreGRID)
Contributing Authors
Marco Aldinucci Department ofComputer Science, University of Pisa, Largo
Bruno Pontecorvo 3, 56127 Pisa, Italy ()
Francoise Andre IRIS A / University of Rennes
1,
Avenue
du
General Leclerc,

35042 Rennes, France ()
Sergio Andreozzi INFN-CNAF, Viale Berti Pichat 6/2, 40126 Bologna, Italy
()
Demetres Antoniades Institute of Computer Science, Foundation for Re-
search and Technology-Hellas, P.O. Box 1385, 71110 Heraklion-Crete, Greece
(danton @ ics.forth.gr)
Elias Athanasopoulos Institute of Computer Science, Foundation for Re-
search and Technology-Hellas, P.O. Box 1385, 71110 Heraklion-Crete, Greece
()
Rosa M. Badia Computer Architecture Department, Universitat Politecnica
de Catalunya, Spain ()
Zoltan Balaton Computer and Automation Research Institute, Hungarian
Academy of Sciences (MTA-SZTAKI), PO.Box 63, 1528 Budapest, Hungary
(balaton @ sztaki.hu)
Francoise Baude INRIA, CNRS-I3S, University of Nice Sophia-Antipolis,
France ()
xii
INTEGRATED RESEARCH IN GRID COMPUTING
Anne Benoit LIP, Ecole Normale Superieure de Lyon, 46 Allee d'ltalie,
69364 Lyon Cedex 07, France ()
Jeremy Buisson IRIS
A
/ University of Rennes 1, Avenue du General Leclerc,
35042 Rennes, France ()
Sonia Campa Department of Computer Science, University of Pisa, Largo
Bruno Pontecorvo 3, 56127 Pisa, Italy ()
Augusto Ciuffoletti INFN-CNAF, Viale Berti Pichat 6/2, 40126 Bologna,
Italy ()
Carmela Comito
DEIS,

University of Calabria, Italy ()
Massimo Coppola ISTI, Area della Ricerca CNR, 56124 Pisa, Italy
(coppola @
di.
unipi. it)
Julita Corbalan Computer Architecture Department, Universitat Politecnica
de Catalunya, Spain ()
Kevin Cristiano Ecole d'Ingenieurs et d'Architectes, 1705 Fribourg,
Switzerland ()
Natalia Currle-Linde High Performance Computing Center (HLRS), Uni-
versity of Stuttgart, Germany ()
Marco Danelutto Department of Computer Science, University of Pisa,
Largo Bruno Pontecorvo 3, 56127 Pisa, Italy ()
Jin Denemark Institute of Computer Science, Masaryk University,
Botanicka 68a, 60200 Brno, Czech Republic ()
Marios Dikaiakos Department of Computer Science, University of Cyprus,
P.O.
Box 537, 1678 Nicosia, Cyprus ()
Piotr Domagalski Poznan Supercomputing and Networking Center,
Noskowskiego 10, 60688 Poznan, Poland ()
Contributing Authors xiii
Jan Diinnweber University of Munster, Department of Mathematics and
Computer Science, Einsteinstrasse 62, 48149 Mtinster, Germany
()
Maciej Dyczkowski Wroclaw Center for Networking and Supercomputing,
Wroclaw University of Technology ()
Dick H.J, Epema Faculty of Electrical Engineering, Mathematics, and Com-
puter Science, Delft University of Technology, Mekelweg 4, 2628 CD, Delft,
The Netherlands ()
Paraskevi Fragopoulou Institute of Computer Science, Foundation for Re-

search and Technology-Hellas, P.O. Box 1385, 71110 Heraklion-Crete, Greece
()
Antonia Ghiselli INFN-CNAF, Viale Berti Pichat 6/2, 40126 Bologna, Italy
(antonia. ghiselli @
cnaf.
infn. it)
Gabor Gombas Computer and Automation Research Institute, Hungarian
Academy of Sciences (MTA-SZTAKI), PO.Box 63, 1528 Budapest, Hungary
(gombasg @ sztaki. hu)
Sergei Gorlatch University of Munster, Department of Mathematics and
Computer Science, Einsteinstrasse 62, 48149 Mtinster, Germany
()
Anastasios Gounaris School of Computer
Science,
University of Manchester,
UK ()
William Groleau Institut National des Sciences Appliquees de Lyon (INSA),
Lyon, France ()
Ralf Gruber Ecole Poly technique Federale de Lausanne, 1015 Lausanne,
Switzerland ()
Francesc Guim Computer Architecture Department, Universitat Politecnica
de Catalunya, Spain ()
xiv
INTEGRATED RESEARCH IN GRID COMPUTING
William Hoarau LRI-CNRS 8623 & INRIA Grand Large, Universite Paris
Sud XI, France ()
Alexandra losup Faculty of Electrical Engineering, Mathematics, and Com-
puter Science, Delft University of Technology, Mekelweg 4, 2628 CD, Delft,
The Netherlands ()
Michal Jankowski Poznan Supercomputing and Networking Center,

Noskowskiego 10, 60688 Poznan, Poland ()
Peter Kacsuk Computer and Automation Research Institute, Hungarian
Academy of Sciences (MTA-SZTAKI), PO.Box 63, 1528 Budapest, Hungary
(kac suk @ sztaki. hu)
Vincent Keller Ecole Poly technique Federale de Lausanne, 1015 Lausanne,
Switzerland (vincent.keller@ epfl.ch)
Thilo Kielmann Dept. of Computer Science, Vrije Universiteit,
Amsterdam, The Netherlands ()
Pierre Kuonen Ecole
d'lngenieurs
et d'Architectes, 1705 Fribourg,
Switzerland ()
Krzysztof Kurowski Poznan Supercomputing and Networking Center,
Noskowskiego 10,60688 Poznan, Poland ()
Agnieszka Kwiecieii Wroclaw Center for Networking and Supercomputing,
Wroclaw University of Technology ()
Jesus Labarta Computer Architecture Department, Universitat Politecnica
de Catalunya, Spain ()
Alexandres Labrinidis Department of Computer Science, University of
Pittsburgh, Pittsburgh 15260, USA ()
Sebastien Lacour IRIS A/INRIA, Campus de Beaulieu,
35042
Rennes Cedex,
France
Contributing Authors
XV
Virginie Legrand INRIA, CNRS-I3S, University of Nice Sophia-Antipolis,
France ()
Robert Lovas Computer and Automation Research Institute, Hungarian
Academy of Sciences (MTA-SZTAKI), RO.Box 63, 1528 Budapest, Hungary

()
Jason Maassen Dept. of Computer Science, Vrije Universiteit,
Amsterdam, The Netherlands ()
Sergio Maffioletti Swiss National Supercomputer Centre, 1015 Manno,
Switzerland ()
Evangeios P. Markatos Institute of Computer Science, Foundation for Re-
search and Technology-Hellas, RO. Box 1385, 71110 Heraklion-Crete, Greece
()
Ludek Matyska Institute of Computer Science, Masaryk University,
Botanicka 68a, 60200 Brno, Czech Republic ()
Norbert Meyer Poznan Supercomputing and Networking Center,
Noskowskiego 10, 60688 Poznan, Poland ()
Jarek Nabrzyski Poznan Supercomputing and Networking Center,
Noskowskiego 10, 60688 Poznan, Poland ()
Nello Nellari Swiss National Supercomputer Centre, 1015 Manno,
Switzerland ()
Rob van Nieuwpoort Dept. of Computer Science, Vrije Universiteit,
Amsterdam, The Netherlands ()
Ariel Oleksiak Poznan Supercomputing and Networking Center,
Noskowskiego 10, 60688 Poznan, Poland ()
Charis Papadakis Institute of Computer Science, Foundation for Research
and Technology-Hellas, P.O. Box 1385, 71110 Heraklion-Crete, Greece
()
xvi
INTEGRATED RESEARCH IN GRID COMPUTING
Nikos Parlavantzas Harrow School of Computer Science, University of
Westminster, HAl 3TP, UK ()
Marcelo Pasin Universidade Federal de Santa Maria, Santa Maria
RS,
Brasil

()
Christian Perez IRISA/INRIA, Campus de Beaulieu, 35042 Rennes Cedex,
France ()
Josep M. Perez Computer Architecture Department, Universitat Politecnica
de Catalunya, Spain ()
Michalis Polychronakis Institute of Computer Science, Foundation for Re-
search and Technology-Hellas, P.O. Box 1385, 71110 Heraklion-Crete, Greece
()
Konstantin Popov Swedish Institute of Computer Science (SICS), Kista,
Sweden ()
Thierry Priol IRISA/INRIA, Campus de Beaulieu, 35042 Rennes Cedex,
France ()
Michael Resch High Performance Computing Center (HLRS), University of
Stuttgart, Germany ()
Miroslav Ruda Institute of Computer Science, Masaryk University,
Botanicka 68a, 60200 Brno, Czech Republic ()
Rizos Sakellariou School of Computer Science, University of Manchester,
UK ()
Marie-Christine Sawley Swiss National Supercomputer Centre,
1015
Manno,
Switzerland ()
Luis Silva Dep. Engenharia Informatica, University of Coimbra, Polo II,
3030 Coimbra, Portugal ()
Contributing Authors
xvii
Gergely Sipos Computer and Automation Research Institute, Hungarian
Academy of Sciences (MTA-SZTAKI),
P.O.Box
63, 1528 Budapest, Hungary

()
Raiil Sirvent Computer Architecture Department, Universitat Politecnica de
Catalunya, Spain ()
Michela Spada Ecole Poly technique Federale de Lausanne, 1015 Lausanne,
Switzerland ()
Domenico Talia DEIS, University of Calabria, Italy ()
Sebastien Tixeuil LRI-CNRS 8623 & INRIA Grand Large, Universite Paris
Sud XI, France ()
Nicola Tonellotto ISTI, Area della Ricerca CNR, 56124 Pisa, Italy
()
Trach-Minh Tran Ecole Poly technique Federale
de
Lausanne,
1015
Lausanne,
Switzerland (trach-minh.tran @ epfl.ch)
Panes Trimintzios European Network and Information Security Agency,
P.O.
Box 1309, 71001 Heraklio, Greece ()
Eleni Tsiakkouri Department of Computer Science, University of Cyprus,
P.O.
Box 537, 1678 Nicosia, Cyprus ()
Vladimir Vlassov Royal Institute of Technology (KTH), Stockholm, Sweden
()
Oliver Waldrich Institute
SCAI,
Fraunhofer Gesellschaft, 53754
St.
Augustin,
Germany (oliver. waeldrich @ scai.fraunhofer.de)

Philipp Wieder Forschungszentrum Jtilich GmbH, 52425 Julich, Germany
()
xviii INTEGRATED RESEARCH IN GRID COMPUTING
Marcin Wojtkiewicz Wroclaw Center for Networking and Supercomputing,
Wroclaw University of Technology ()
Pawel Wolniewicz Poznan Supercomputing and Networking Center,
Noskowskiego 10, 60688 Poznan, Poland ()
Gosia Wrzesinska Dept. of Computer Science, Vrije Universiteit,
Amsterdam, The Netherlands ()
Ramin Yahyapour Robotics Research Institute, University of Dortmund,
44221 Dortmund, Germany ()
Henan Zhao School of Computer Science, University of Manchester, UK
()
Wolfgang Ziegler Institute
SCAI,
Fraunhofer Gesellschaft, 53754
St.
Augustin,
Germany ()
Corrado Zoccolo Department of Computer Science, University of Pisa,
Largo Bruno Pontecorvo 3, 56127 Pisa, Italy ()
DATA INTEGRATION AND
QUERY REFORMULATION IN
SERVICE-BASED GRIDS
Carmela Comito and Domenico Talia
DEIS,
University of
Calabria,
Italy

Anastasios Gounaris and Rizos Sakellariou
School of Computer
Science,
University of Manchester, UK
gounaris @cs.man.ac.ul<

Abstract
This paper describes the XMAP data integration framework and query refor-
mulation algorithm, provides insights into the performance of the algorithm, and
about its use in implementing query processing services. Here we propose an ap-
proach for data integration-enabled distributed query processing on Grids by em-
bedding the XMAP reformulation algorithm within the OGSA-DQP distributed
query processor. To this aim we exploit the OGSA-DQP XML representation
of relational schemas by applying the XMAP algorithm on them. Moreover, we
introduce a technique to rewrite an XPath query into an equivalent OQL one.
Finally, the paper presents a roadmap for the integration system implementation
aiming at constructing an extended set of services that will allow users to submit
queries over a single database and receive the results from multiple databases
that are semantically con'elated with the former one.
Keywords: XML databases, semantic data integration, schema mappings, distributed query
processing, Grid services.
2
INTEGRATED RESEARCH IN GRID COMPUTING
1.
Introduction
The Grid offers new opportunities and raises new challenges in data manage-
ment that originate from the large scale, dynamic, autonomous, and distributed
nature of data sources. A Grid can include related data resources maintained
in different syntaxes, managed by different software systems, and accessible

through different protocols and interfaces. Due to this diversity in data re-
sources, one of the most demanding issues in managing data on Grids is recon-
ciliation of data heterogeneity [11]. Therefore, in order to provide facilities for
addressing requests over multiple heterogeneous data sources, it is necessary
to provide data integration models and mechanisms.
Data integration is the flexible and managed federation, analysis, and pro-
cessing of data from different distributed sources. In particular, the increase in
availability of web-based data sources has led to new challenges in data integra-
tion systems for obtaining decentralized, wide-scale sharing of
data,
preserving
semantics. These new needs in data integration systems are also felt in Grid
settings. In a Grid, a centralized structure for coordinating all the nodes is not
efficient because it can represent a bottleneck and, more importantly, it cannot
accommodate the dynamic and distributed nature of Grid resources.
The Grid community is devoting great attention toward the management of
structured and semi-structured data such as relational and XML data. Two
significant examples of such efforts are the OGSA Data Access and Integra-
tion (OGSA-DAI) [3] and the OGSA Distributed Query Processor (OGSA-
DQP) [2] projects. However, till today only few projects (e.g., [8, 6]) actually
meet schema-integration requirements necessary for establishing semantic con-
nections among heterogeneous data sources.
For these reasons, we propose the use of the XMAP framework [9] for
integrating heterogeneous data sources distributed over a Grid. By means of
this framework, we aim at developing a decentralized network of semantically
related schemas that enables the formulation of distributed queries over het-
erogeneous data sources. We designed a method to combine and query XML
documents through a decentralized point-to-point mediation process among
the different data sources based on schema mappings. We offer a decentral-
ized service-based architecture that exposes this XML integration formalism

as an e-Service. The infrastructure proposed exploits the middleware provided
by OGSA-DQP and OGSA-DAI, building on top of them schema-integration
services.
The remainder of the paper is organized as follows. Section
2
presents a short
analysis of data integration systems focusing on specific issues related to Grids.
Section
3
presents the XMAP integration framework; the underlying integration
model and
the
XMAP query reformulation algorithm
are
described. The OGSA-
DQP and OGSA-DAI existing query processing services are outlined in Section
Data integration and query reformulation in service-based Grids 3
4.
Section 5 presents an example of applying the XMAP algorithm to OGSA-
DQP,
whereas Section 6 introduces the approach proposed to rewrite an XPath
query into an equivalent OQL one. Finally, Section 8 concludes the paper.
2.
Data Integration in Grids
The goal of a data integration system is to combine heterogeneous data
residing at different sites by providing a unified view of this data. The two
main approaches
to
data integration
are

federated database management systems
(FDBMSs) and traditional mediator/wrapper-based integration systems.
A federated database management system (FDBMS) [19] is a collection of
cooperating but autonomous component database systems (DBSs). The DBMS
of
a
component DBS, or component DBMS, can be a centralized or distributed
DBMS or another FDBMS. The component DBMSs can differ in different
aspects such as data models, query languages, and transaction management
capabilities.
Traditional data integration systems [17] are characterized by an architecture
based on one or more mediated schemas and a set of sources. Each source
contains data, while every mediated schema provides a reconciled, integrated,
and virtual view of
the
underlying sources. Moreover, the system includes a set
of source descriptions that provide semantic mappings between the relations in
the source schemas and the relations in the mediated schemas [18] .
Data integration on Grids presents a twofold characterization:
1 data integration is a key issue for exploiting the availability of large,
heterogeneous, distributed and highly dynamic data volumes on Grids;
2 integration formalisms can benefit from an OGS A-based Grid infrastruc-
ture,
since it facilitates dynamic discovery, allocation, access, and use
of both data sources and computational resources, as required to support
computationally demanding database operations such as query reformu-
lation, compilation and evaluation.
Data integration on Grids has to deal with unpredictable, highly dynamic data
volumes provided by unpredictable membership of nodes that happen to be
participating at any given time. So, traditional approaches to data integration,

such as FDBMS [19] and the use of mediator/wrapper middleware [18] , are
not suitable in Grid settings.
The federation approach is a rather rigid configuration where resources al-
location is static and optimization cannot take advantage of evolving circum-
stances in the execution environment. The design of mediator/wrapper inte-
gration systems must be done globally and the coordination of mediators has
been done by a central administrator which is an obstacle to the exploitation
of evolving characteristics of dynamic environments. As a consequence, data