Abstract: The last version of the Globus Toolkit includes both pre-WS and WS GRAM services to submit, monitor, and control jobs on remote Grid resources. In the medium term and until a full transition is accomplished, both pre-WS and WS GRAM services will coexist in Grid infrastructures. In this paper, we describe the modular architecture of the GridWay meta-scheduler, which allows the simultaneous and coordinated use of pre-WS and WS GRAM services and, therefore, makes easy the transition to a Web Service implementation of the Globus components. Such functionality is demonstrated on a infrastructure that comprises resources from a research testbed, based on the Globus Toolkit 4.0, and the EGEE production infrastructure, based on the LCG middleware. The Web Service implementation of Globus components has been optimized for flexibility, stability and scalability. However, part of the Grid community is still reluctant to transition to the Web Service model due mainly to its supposed lower performance. We demonstrate that WS GRAM achieves a performance comparable to that of pre-WS GRAM. (c) 2006 Elsevier B.V. All rights reserved.

Abstract: Since the late 1990s, we have witnessed an extraordinary development of Grid technologies. Nowadays, different Grids are being deployed within the context of a growing number of national and transnational research projects. However. the coexistence of those different infrastructures involves two challenging issues, namely: (i) simultaneous and coordinated use of resources from different Grids. from the end user perspective; and (ii) the simultaneous contribution of resources to different Grids. from the resource owner perspective. In this paper, we demonstrate that a decentralized and "end-to-end" scheduling and execution system can efficiently interoperate different Grids. In particular, we evaluate the coordinated use of the EGEE and IRISGrid testbeds in the execution of a Bioinformatics application. Results show the feasibility of building loosely coupled computational Grid environments only based on Globus services, while obtaining non-trivial levels of quality of service, in terms of performance and reliability. Such approach allows a straightforward resource sharing since the resources are accessed by using de facto standard protocols and interfaces. (c) 2006 Elsevier Inc. All rights reserved.

Abstract: Metascheduling is a key functionality of the grid middleware in order to achieve a reasonable degree of performance and reliability, given the changing conditions of the computing environment. In this contribution a comparative analysis between two major grid scheduling philosophies is shown: a semi-centralized approach, represented by the EGEE Workload Management System, and a fully distributed approach, represented by the Grid Way Metascheduler. This comparative is both theoretical, through a functionality checklist, and experimental, through the execution of a fusion plasma application on the EGEE infrastructure.

Abstract: This paper describes the execution of a Bioinformatics application over a highly distributed and heterogeneous testbed. This testbed is composed of resources devoted to EGEE and IRISGrid projects and has been integrated by taking advantage of the modular, decentralized and "end-to-end" architecture of the Grid Way framework. Results show the feasibility of building loosely-coupled Grid environments based only on Globus services, while obtaining non trivial levels of quality of service. Such approach allows a straightforward resource sharing as the resources are accessed by using de facto standard protocols and interfaces.

Abstract: Grids constitute a promising platform to execute loosely coupled, high-throughput parameter sweep applications, which arise naturally in many scientific and engineering fields like bio-informatics, computational fluid dynamics, particle physics, etc. In spite of the simple computational structure of these applications, its efficient execution and scheduling are challenging because of the dynamic and heterogeneous nature of Grids. In this work, we propose a benchmarking methodology to analyze the performance of computational Grids in the execution of high throughput computing applications, that combines: (i) a representative benchmark included in the NAS Grid Benchmark suite; (ii) a performance model that provides a way to parametrize and compare different Grids; and (iii) a set of application-level performance metrics to analyze and predict the performance of this kind of applications. The benchmarking methodology will be applied to the performance analysis of a Globus-based research testbed that spans heterogeneous resources in five institutions. (c) 2005 Elsevier B.V. All rights reserved.

Abstract: Reliability, in terms of Grid component fault tolerance and minimum quality of service, is an important aspect that has to be addressed to foster Grid technology adoption. Software reliability is critically important in today's integrated and distributed systems, as is often the weak link in system performance. In general, reliability is difficult to measure, and specially in Grid environments, where evaluation methodologies are novel and controversial matters. This paper describes a straightforward procedure to analyze the reliability of computational grids from the viewpoint of an end user. The procedure is illustrated in the evaluation of a research Grid infrastructure based on Globus basic services and the GridWay meta-scheduler. The GridWay support for fault tolerance is also demonstrated in a production-level environment. Results show that GridWay is a reliable workload management tool for dynamic and faulty Grid environments. Transparently to the end user, GridWay is able to detect and recover from any of the Grid element failure, outage and saturation conditions specified by the reliability analysis procedure. (c) 2006 Elsevier B.V. All rights reserved.

Abstract: The large number of protein sequences, provided by genomic projects at an increasing pace, constitutes a challenge for large scale computational studies of protein structure and thermodynamics. Grid technology is very suitable to face this challenge, since it provides a way to access the resources needed in compute and data intensive applications. In this paper, we show the procedure to adapt to the Grid an algorithm for the prediction of protein thermodynamics, using the GridWay tool. GridWay allows the resolution of large computational experiments by reacting to events dynamically generated by both the Grid and the application.

Abstract: Genetic algorithms (GAs) are stochastic search methods that have been successfully applied in many search, optimization, and machine learning problems. Their parallel counterpart (PGA, parallel genetic algorithms) offers many advantages over the traditional GAs, such as speed, ability to search on a larger search space, and less likely to run into a local optimum. With the advent of Grid computing, the computational power that can be deliver to the applications have substantially increased, and so PGAs can potentially benefit from this new Grid technologies. However, because of the dynamic and heterogeneous nature of Grid environments, the implementation and execution of PGAs in a Grid involve challenging issues. This paper discusses the distribution of a PGA across the Grid using the DRMAA standard API and the Grid Way framework. The efficiency and reliability of this schema to solve the One Max problem is analyzed in a globus-based research testbed.

Abstract: The efficient usage of current emerging Grid infrastructures can only be attained by defining a standard methodology for its evaluation. This methodology should include an appropriate set of criteria and metrics, and a suitable family of Grid benchmarks, reflecting representative workloads, to evaluate such criteria and metrics. The establishment of this methodology would be useful to validate the middleware, to adjust its components and to estimate the achieved quality of service.

Abstract: The coexistence of different Grid infrastructures and the advent of Grid services based on Web Services opens an interesting debate about the coordinated harnessing of resources based on different middleware implementations and even different Grid service technologies. In this paper, we present the loosely-coupled architecture of Grid Way, which allows the coordinated use of different Grid infrastructures, although based on different Grid middlewares and services, as well as a straightforward resource sharing. This architecture eases the gradual migration from pre-WS Grid services to WS ones, and even, the long-term coexistence of both. We demonstrate its suitability with the evaluation of the coordinated use of two Grid infrastructures: a research testbed based on Globus WS Grid services, and a production testbed based on Globus pre-WS Grid services, as part of the LCG middleware.

Abstract: Grids are by nature highly dynamic and heterogeneous environments, and this is specially the case for the performance of the interconnection links between grid resources. Therefore, grid resource selection should take into account the proximity of the computational resources to the needed data in order to reduce the cost of file staging. This fact is specially relevant in the case of adaptive job execution, since job migration requires the transfer of large restart files between the compute hosts. In this paper, we discuss the extension of the Grid Way framework to also consider dynamic resource proximity to select grid resources, and to decide if job migration is feasible and worthwhile. The benefits of the new resource selector will be demonstrated for the adaptive execution of a computational fluid dynamics (CFD) code.

Abstract: The large number of protein sequences, provided by genomic projects at an increasing pace, constitutes a challenge for large scale computational studies of protein structure and thermodynamics. Grid technology is very suitable to face this challenge, since it provides a way to access the resources needed in compute and data intensive applications. In this work, we concentrate on the grid-aware implementation of a protein structure prediction algorithm.

Abstract: Grids offer a dramatic increase in the number of available processing and storing resources that can be delivered to applications. However, efficient job submission and management continue being far from accessible to ordinary scientists and engineers due to their dynamic and complex nature. This paper describes a new Globus based framework that allows an easier and more efficient execution of jobs in a 'submit and forget' fashion. The framework automatically performs the steps involved in job submission and also watches over its efficient execution. In order to obtain a reasonable degree of performance, job execution is adapted to dynamic resource conditions and application demands. Adaptation is achieved by supporting automatic application migration following performance degradation, 'better' resource discovery, requirement change, owner decision or remote resource failure. The framework is currently functional on any Grid testbed based on Globus because it does not require new system software to be installed in the resources. The paper also includes practical experiences of the behavior of our framework on the TRGP and UCM-CAB testbeds. Copyright (C) 2004 John Wiley Sons, Ltd.

Abstract: Grids constitute a promising platform to execute loosely coupled applications, which arise naturally in many scientific and engineering fields like bioinformatics, computational fluid dynamics, particle physics, etc. In this paper, we describe our experiences in porting three scientific production codes to the Grid. Those codes follow typical computational models, namely: embarrassingly distributed and master-worker. In spite of their relatively simple computational structure, consisting of many "independent" tasks, their reliable and efficient execution on computational Grids involves several issues, due to both the dynamic nature of the Grid itself and the execution and programming requirements of the applications. The applications have been developed by using the DRMAA (Distributed Resource Management Application API) interface. DRMAA routines are supported by the functionality offered by the Grid Way framework, that provides the runtime mechanisms needed for transparently executing jobs on a dynamic Grid environment. The experiments have been performed on Globus-based research testbeds that span heterogeneous resources in different institutions.

Abstract: In spite of the great research effort made in Grid technology in the last years, application development and execution in the Grid continue requiring a high level of expertise due to its complex and dynamic nature. The Distributed Resource Management Application API (DRMAA) has been proposed to aid the rapid development and distribution of applications across the Grid. In this paper we present the first implementation of the DRMAA standard on a Globus-based testbed, and show its suitability to express typical scientific applications. The DRMAA routines have been implemented using the functionality provided by the Grid Way framework.

Abstract: Marine-target impact cratering simulation plays an important role in the study of past martian seas. In order to develop profiles for future exploration missions and to understand the morphologies for future investigations, a large number of simulations have to be done. This paper presents some experimental results obtained with the execution of impact cratering simulations in a Globus-based Grid environment through the GridWay framework. Some performance metrics are presented to demonstrate the suitability of this platform for running High Throughput Computing applications.

Abstract: Grids provide a way to access the resources needed to execute the compute and data intensive applications required in the Bioinformatics field. However, in spite of the great research effort made in the last years, application development and execution in the Grid continue requiring a high level of expertise due to its heterogeneous and dynamic nature. In this paper, we show the procedure to adapt an existing Bioinformatics application to the Grid using the GridWay tool. The GridWay allows the efficient resolution of large computational experiments by reacting automatically to Grid- and application-generated dynamic events.

Abstract: The ability to migrate running applications among different grid resources is generally accepted as the solution to adapt to dynamic resource load, availability and cost. In this paper we focus on opportunistic migration when a new resource becomes available in the Grid. In this situation the performance of the new host, the remaining execution time of the application, and also the proximity of the new resource to the needed data, become critical factors to decide if job migration is feasible and worthwhile. We discuss the extension of the GridWay framework to consider all the previous factors in the resource selection and migration stages in order to improve response times of individual applications. The benefits of the new resource selector will be demonstrated for the execution of a computational fluid dynamics (CFD) code.

Abstract: The aim of this paper is to study the influence of processor mapping on message passing performance of two different parallel computers: the Cray T3E and the SGI Origin 2000. For this purpose, we have first designed an experiment where processors are paired off in a random manner and messages are exchanged between them. In view of the results of this experiment, it is obvious that the physical placement must be accounted for. Consequently, a mapping algorithm for the Cray T3E, suited cartesian topologies is studied. We conclude by making comparisons between our T3E algorithm, the MPI default mapping and another algorithm proposed by Muller and Resch in [9].