ABSTRACT ACCEPTED

Authors = "G. Avellino, S. Barale, B. Cantalupo, D. Colling, F. Giacomini, A. Gianelle,
A. Guarise, A. Krenek, D. Kouril, A. Maraschini, L. Matyska, M. Mezzadri, S. Monforte, M. Mulac,
F. Pacini, M. Pappalardo, R. Peluso, J. Pospisil, F. Prelz, E. Ronchieri, M. Ruda, L. Salconi,
Z. Salvet, M. Sgaravatto, J. Sitera,
A. Terracina, M. Vocu, A. Werbrouck"

Presentation Type = "Oral Presentation"

Abstract = "Application users have now been experiencing for about a year now with 
the standardized resource brokering services provided by the 'workload management' package of the EU DataGrid project (WP1).
Understanding, shaping and pushing the limits of the system has provided valuable feedback on both its design and implementation. A digest of the lessons, and "better practices", that  were learned, and that were applied towards the second major release of the software, is given."

ABSTRACT ACCEPTED


Authors: G. Avellino, S. Barale, B. Cantalupo, D. Colling, F. Giacomini, A. Gianelle,
A. Guarise, A. Krenek, D. Kouril, A. Maraschini, L. Matyska, M. Mezzadri, S. Monforte, M. Mulac, F. Pacini, M. Pappalardo, R. Peluso, J. Pospisil, F. Prelz, E. Ronchieri, M. Ruda, L. Salconi,
Z. Salvet, M. Sgaravatto, J. Sitera, A. Terracina, M. Vocu, A. Werbrouck

Abstract: In the first phase of the European DataGrid project, the 'workload management'
package (WP1) implemented a working prototype, providing users with an
environment allowing to define and submit jobs to the Grid, and able to find
and use the ``best'' resources for these jobs. Application users have now
been experiencing for about a year now with this first release of the
workload management system.
The experiences acquired, the feedback received by the user and the
need to plug new components implementing new functionalities, triggered
an update of the existing architecture.
A description of this revised and complemented 
workload management system is
given.

ABSTRACT ACCEPTED


Authors: Heinz Stockinger, Flavia Donno, Erwin Laure, Shahzad Muzaffar, Giuseppe Andronico, Peter Kunszt, Paul Millar

In the first phase of the EU DataGrid (EDG) project, a Data Management System has been implemented and provided for deployment. The components of the current EDG Testbed are: a prototype of a Replica Manager Service built around the basic services provided by Globus, a centralized Replica Catalogue to store information about physical locations of files, and the Grid Data Mirroring Package (GDMP) that is widely used in various HEP collaborations in Europe and the US for data mirroring. During this year these services have been refined and made more robust so that they are fit to be used in a pre-production environment. Application users have been using this first release of the Data Management Services for more than a year. In the paper we present the components and their interaction, our implementation and experience as well as the feedback received from our user communities. We have resolved not only issues regarding integration with other EDG service components but also many of the interoperability issues with components of our partner projects in Europe and the U.S. The paper concludes with the basic lessons learned during this operation. These conclusions provide the motivation for the architecture of the next generation of Data Management Services that will be deployed in EDG during 2003.

AUTHORS
         I.Augustin,R.Barbera,J.J.Blaising,K.Bos,D.Boutigny,S.Burke,P.Capiluppi,F.Carminati,
         P.Cerello,J.Closier,F.Harris,E.van.Herwijnen,L.Perini,M.Reale,S.Resconi,
         A.Sciaba,M.Sitta,O.Smirnova,J.Templon,A.Tsaregorodtsev

EDITOR  F.Harris           PRESENTER   S.Burke


ABSTRACT ACCEPTED


Workpackage 8(WP8) of the European Datagrid (EDG) project was formed in Jan 2001
with representatives from the 4 LHC experiments, and with experiment independent
people (EIPs) from  5 of the 6 main EDG partners. In September 2002 WP8 has been
strengthened by the addition of effort from Babar and D0, and more recently NA48.
The original mandate of WP8  was, following the definition of short and long term requirements,
to port experiment software to the EDG middleware and testbed environment. A major additional
activity has been testing the basic functionality and performance of this environment.

The paper reviews the experiences and the evaluations in the areas of
job submission, data management, mass storage handling, information systems and monitoring.
It also comments on the problems of remote debugging, the portability of code,  and scaling
problems with increasing numbers of jobs, sites and nodes. Reference is made to the 
pioneeering work of Atlas and CMS in integrating the use of EDG Testbed-1 into their
data challenges. A forward look is made to essential software developments within EDG and to
the necessary cooperation between EDG and LCG for the LCG prototype-1 due in mid 2003.

ABSTRACT ACCEPTED


Authors = 
R. Alfieri, F. Bonnassieux, P. Broadfoot, R. Cecchini, V. Ciaschini, L. Cornwall, L. dell'Agnello, A. Frohner, A. Gianoli, D. Groep, J. Jensen, D. Kelsey, D. Kouril, G. Lowe, A. McNab, O. Mulmo, F. Spataro

Oral Presentation
Luca dell'Agnello (INFN)

Category 1 - Grid Architecture, Infrastructure and middleware

One of the fundamental concepts in Grid computing is the creation of
Virtual Organisations (VOs): a set of resource consumers and providers
that join forces to solve a common problem. The collaborations that
formed around, e.g., the LHC experiments, are typical examples of
Virtual Organisations. To date, Grid computing has been applied on a
relatively small scale, linking dozens of users to a dozen resources,
and management of these VOs was a largely manual operation. With the
advance of large collaboration, linking more than 10000 users with a
thousand sites in 150 counties, a comprehensive, automated management
system is required. It should be simple enough not to deter users,
while at the same time ensuring local site autonomy.
The VO Management Service (VOMS), developed by the EU DataGrid
project, is a strongly secured system for managing users and resources
in virtual organisations. It extends the existing Grid Security
Infrastructure architecture with embedded VO affiliation assertions
that can be independently verified by all VO members and resource
providers. Within the EU DataGrid project, Grid services for job
submission, file- and database access are being equipped with fine-
grained authorization systems that take VO membership into account.
These also give resource owners the ability to ensure site security
and enforce local access policies.

This talk will describe the EU DataGrid security architecture, the VO
membership service and the local site enforcement mechanisms LCAS,
LCMAPS and the Spitfire TrustManager.

Abstract accepted

ABSTRACT ACCEPTED

Authors:

Rob Byrom, Brian Coghlan, Andrew W Cooke, Roney
Cordenonsi, Linda Cornwall, Ari Datta, Abdeslem Djaoui, Laurence
Field, Steve Fisher, Steve Hicks, Stuart Kenny, James Magowan, Werner
Nutt, David O'Callaghan, Manfred Oevers, Norbert Podhorszki, John
Ryan, Manish Soni, Paul Taylor, Antony J. Wilson and Xiaomei Zhu

Abstract:

We describe R-GMA (Relational Grid Monitoring Architecture) which is
being developed within the European DataGrid Project as an Grid
Information and Monitoring System. Is is based on the GMA from GGF,
which is a simple Consumer-Producer model. The special strength of
this implementation comes from the power of the relational model. We
offer a global view of the information as if each VO had one large
relational database.  We provide a number of different Producer types
with different characteristics; for example some support streaming of
information. We also provide combined Consumer/Producers, which are
able to combine information and republish it. At the heart of the
system is the mediator, which for any query is able to find and
connect to the best Producers to do the job.  We are able to invoke
MDS info-provider scripts and publish the resulting information via
R-GMA in addition to having some of our own sensors.  APIs are
available which allow the user to deploy monitoring and information
services for any application that may be needed in the future.  We
have used it both for information about the grid (primarily to find
what services are available at any one time) and for application
monitoring.  R-GMA has been deployed in Grid testbeds in the United
Kingdom and Italy, and we describe the results and experiences of this
deployment.

Note:

All names are subject to change

Title: EU DataGRID testbed management and support at CERN

ABSTRACT ACCEPTED

Authors: Emanuele Leonardi, Markus Schulz et al.

Text: In this talk we report on the first two years of running of the CERN testbed site for the EU DataGRID project. The site consists of about 120 dual-processor PCs distributed over several testbeds used for different purposes: software development, system integration, and application tests. Activities at the site included test productions of MonteCarlo data for LHC experiments, tutorials and demonstrations of GRID technologies, and support for individual user analysis. The talk will focus on nodes installation and configuration techniques, services management, user support in a GRID-ified environment, and will include considerations on scalability and security issues and comparisons with "traditional" production systems, as seen from the administrator point of view.

Type: Oral Presentation

Category: Category 2: HENP Grid Applications, Testbeds, & Demonstrations

Presenter: Emanuele Leonardi 

Presenter Institution: CERN

ACCEPTED

Authors:
James Casey, Diana Bosio, Akos Frohner, Leanne Guy, Joni Hahkala, Peter Kunszt, Erwin Laure, Sophie Lemaitre, Levi Lucio, Gavin McCance, Olle Mulmo, Heinz Stockinger, Kurt Stockinger, Giuseppe Andronico, William Bell, David Cameron, Ruben Carvajal-Schiaffino, Federico DiCarlo, Andrea Domenici, Wolfgang Hoschek, Niklas Karlsson, Paul Millar, Ville Nenonen, Livio Salconi, Mika Silander, Gian-Luca Volpato, Floriano Zini

Text:
We describe the architecture and initial implementation of the next-generation of Grid Data Management Middleware in the EU DataGrid (EDG) project.

The new architecture stems out of our experience and the users 
requirements gathered during the two years of running our initial set of 
Grid Data Management Services. All of our new services are based on the 
Web Service technology paradigm, very much in line with the emerging 
Open Grid Services Architecture (OGSA). We have modularized our 
components and invested a great amount of effort towards a secure, 
extensible and robust service, starting from the design but also using a 
streamlined build and testing framework.

Our service components are: Replica Location Service, Replica Metadata 
Service, Replica Optimization Service, Replica Subscription and 
high-level replica management. The service security infrastructure is 
fully GSI-enabled, hence compatible with the existing Globus Toolkit 
2-based services; moreover, it allows for fine-grained authorization 
mechanisms that can be adjusted depending on the service semantics.

Type:
Oral Presentation

Category:
Category 1: Grid Architecture, Infrastructure, & Middleware

ACCEPTED

*Authors:* Diana Bosio, James Casey, Dirk D"ullmann, Maria Girone,
Leanne Guy, Peter Kunszt, Zhen Xie

*Text:* The POOL project has a need for both local file catalogs and
Grid accessible catalogs in order to locate replicas of data both
locally and on the Grid. In this paper we present the requirements,
design and architecture of the POOL File Catalog and its grid aware
implementation. This is based on Replica Location and Metadata Services
produced by the EU DataGrid and Globus projects. We look at the specific
requirements for the catalog subsystem of POOL, and the choices they
imply in terms of software solution. We describe the architecture of the
catalog subsystem, based on Web Service and Relational Database
technologies and comment on our experiences in deploying this type of
system. Finally, we examine our choices in determining the underlying
components, in terms of the quality of service that we wish to provide.
This includes not only the choice of hardware platform, but also the
software platform, in particular the RDBMS and application server. We
contrast the experience of deploying on both an open-source solution,
i.e. Jakarta Tomcat and MySQL, as well as on a commercial one, i.e.
Oracle 9i and 9iAS.

*Type:* Oral Presentation

*Category:* Category 2: HENP Grid Applications, Testbeds, & Demonstrations