The first deployment of workload management services on the EU DataGrid Testbed: feedback on design and implementation.

• Francesco Prelz (Unknown)

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."

Towards automation of computing fabrics using tools from the fabric managmenet workpackage of the EU DataGrid project

• Olof Barring (CERN)

<B>ABSTRACT ACCEPTED</B> <UL> <B>Authors:</B> <UL> <LI>CERN: Olof Bärring, Maite Barroso Lopez, German Cancio, Sylvain Chapeland, Lionel Cons, Piotr Poznañski, Philippe Defert, Jan Iven, Thorsten Kleinwort, Bernd Panzer-Steindel, Jaroslaw Polok, Catherine Rafflin, Alan Silverman, Tim Smith, Jan Van Eldik</LI> <LI>INFN: Massimo Biasotto, Andrea Chierici, Luca Dellagnello, Gaetano Maron, Michele Michelotto, Cristine Aiftimiei, Marco Serra, Enrico Ferro</LI> <LI>ZIB: Thomas Röblitz, Florian Schintke</LI> <LI>EVG UNI HEI: Lord Hess, Volker Lindenstruth, Frank Pister, Timm Morten Steinbeck</LI> <LI>NIKHEF/FOM: David Groep, Martijn Steenbakkers</LI> <LI>PPARC: Paul Anderson, Tim Colles, Alexander Holt, Alastair Scobie, Michael George <P> <B>Abstract:</B> The EU DataGrid project workpackage 4 has as an objective to provide the necessary tools for automating the management of medium size to very large computing fabrics. At the end of the second project year subsystems for centralised configuration management (presented at LISA'02) and performance/exception monitoring have been delivered. This will soon be augumented with a subsystem for node installation and service configuration, which is based on existing widely used standards where available (e.g. rpm, kickstart, init.d scripts) and clean interfaces to OS dependent components (e.g. base installation and service management). The three subsystems together allow for centralised management of very large computer farms. Finally, a fault tolerance system is being developed for tying together the above subsystems to form a complet framework for automated enterprise computing management by 3Q03. All software developed is open source covered by the EU DataGrid project license agreements. <P> This article describes the architecture behind the designed fabric management system and the status of the different developments. It also covers the experience with an existing tool for automated configuration and installation that have been adapted and used from the beginning to manage the EU DataGrid testbed, which is now used for LHC data challenges.

The EU DataGrid Workload Management System: towards the second major release

• Massimo Sgaravatto (INFN PADOVA)

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.

Grid Data Management in Action: Experience in Running and Supporting Data Management Services in the EU DataGrid Project.

• Flavia Donno (CERN AND INFN)

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.

Grid Applications,Testbeds and demonstrations - HEP Applications evaluation of the EDG Testbed and Middleware

• Olof Barring (CERN)

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.

R-GMA: First results after deployment

• Steve Fisher (RAL)

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

EU DataGRID testbed management and support at CERN

• Olof Barring (CERN)

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

Next-Generation EU DataGrid Data Management Services.

• James Casey (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

Grid Accessible File and Metadata Catalogs in the LCG POOL Project

• Bruno Puccio (CERN)

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