Made available in DSpace on 2014-07-29T14:57:51Z (GMT). No. of bitstreams: 1 dissertacaoRaphaelGomes.pdf: 1868602 bytes, checksum: 672048d88b1eb45c74f869fe6794261f (MD5) Previous issue date: 2009-04-13 / Scavenging grid computing environments are being even more used as an alternative aside traditional processing categories. This kind of system is characterized by using of communal and shared workstations and its main concern is the guarantee of performance for the user who gives its resources to the grid. This concern, and because it is an extremaly dynamic environment, causes the adoption of treatments limited to best-effort. However, this effort hardly ever is fulfilled in an optimal way due the fact that are usually used treatments which doesn t create an effective managing of resources, doing the competitiveness of the grid applications with the local applications always penalize the grid s tasks, lowing the its performance. This work presents an architecture for the performance improvement to the grid applications, which takes into account using profiles of resources for locals applications, trying to identify when it is temporary and avoid taking actions as grid tasks migration. As alternative way or additional is proposed the using of adaptive technics who focus allowing the grade tasks maintenance on the actual node or the system improvement. The architecture was partially implemented on InteGrade middleware. We describe the developed implementation an its evaluation / Ambientes de grade oportunista vêm sendo cada vez mais usados como alternativa a categorias tradicionais de processamento. Esse tipo de sistema se caracteriza pela utilização de estações de trabalho comuns e compartilhadas e tem como principal preocupação a garantia de desempenho para o usuário que cede seus recursos à grade. Essa preocupação, aliada ao fato deste ser um ambiente computacional extremamente dinâmico, faz com que os tratamentos adotados no sistema sejam limitados ao melhor-esforço. Contudo, esse esforço nem sempre é cumprido da melhor forma possível pois geralmente são adotados tratamentos que não realizam um gerenciamento eficiente dos recursos, fazendo com que a competitividade das aplicações da grade com as aplicações locais penalize sempre as tarefas da grade, diminuindo o desempenho destas. Neste trabalho é apresentada uma arquitetura para melhoria de desempenho para aplicações da grade, que leva em consideração perfis de utilização de recursos por aplicações locais, tentando identificar quando isso é algo passageiro e evitar a tomada de ações como migração de tarefas da grade. Como medida alternativa ou adicional é proposto o uso de técnicas adaptativas que visam permitir a manutenção das tarefas da grade no nó atual ou melhorar o sistema. A arquitetura foi parcialmente implementada no middleware InteGrade. Descrevemos a implementação desenvolvida e sua avaliação

Made available in DSpace on 2016-08-17T14:53:06Z (GMT). No. of bitstreams: 1 Diego_Souza_Gomes.pdf: 2086164 bytes, checksum: d2a9f245dd16522cb21c6fcb10bb8447 (MD5) Previous issue date: 2009-12-07 / Mobile computing technologies and wireless networks have evolved very quickly. Nowadays, many of the portable devices have significant processing power, storage and communication capacity. At the same time, grid computing technology became an important tool for collaborative work among users and organizations enabling the sharing of computing resources and services through multiple administrative domains. Due to the popularity of mobile computing, users of handheld devices form an important and new segment of the computing grid, assuming both the role of consumers and providers of resources and services. The integration of these two categories of distributed systems aims to extend the capabilities of portable devices, enabling access to a infrastructure of shared resources, as well as provide faster and easier means for grid services users to access information produced by these systems at any time and place. This work describes the MInteGrade (Mobile InteGrade), a software infrastructure that provides access to the InteGrade grid middleware from mobile devices connected through IEEE 802.11 wireless networks in infrastructure mode, and Bluetooth networks in Ad hoc mode. Through this mechanism for access to a computational grid, mobile clients can request the execution of applications on the grid, monitor the execution of applications and view the computation results already completed. The MInteGrade was designed so as to consider the dynamics of wireless networks. Specifically, it supports disconnections of devices and dynamic changes of the Ad hoc network topology, aswell as content adaptation of the output of the grid computations. / Tecnologias de computação móvel e de redes sem fio têm evoluído muito rapidamente, de forma que muitos dos dispositivos portáteis possuem atualmente considerável capacidade de processamento, armazenamento e comunicação. Paralelo a isso, a tecnologia de grades computacionais se consolidou como um ferramental importante para o trabalho colaborativo entre usuários e organizações, através do compartilhamento de recursos e serviços computacionais entre múltiplos domínios administrativos. Devido a esta popularização da computação móvel, usuários de dispositivos portáteis formam um importante e novo segmento da computação em grade, assumindo tanto o papel de usuários como também de provedores de recursos e serviços. A integração dessas duas categorias de sistemas distribuídos, objetiva estender as capacidades dos dispositivos computacionais móveis através do acesso a uma infra-estrutura de recursos compartilhados, além de fornecer aos usuários de serviços de grade meios mais rápidos e fáceis de acesso as informações produzidas por estes sistemas em qualquer hora e lugar. Este trabalho descreve oMInteGrade (Mobile InteGrade), uma infra-estrutura de software para acesso aos serviços do middleware de grade InteGrade a partir de dispositivos móveis conectados através de redes sem fio IEEE 802.11 em modo infraestruturado e redes bluetooth em modo Ad hoc. Através deste mecanismo para acesso a uma grade de computadores, clientes móveis podem solicitar a execução de aplicações na grade, realizar o acompanhamento da execução das aplicações e visualizar o resultado de computações já concluídas. O MInteGrade foi projetado para levar em consideração o dinamismo das redes sem fio, provendo suporte a períodos de desconexão e variações na topologia das redes Ad hoc, bem como a adaptação de conteúdo dos resultados das computações realizadas pela grade.

Orientador: Edmundo Roberto Mauro Madeira / Dissertação (mestrado) ¿ Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-24T01:33:20Z (GMT). No. of bitstreams: 1 Silva_GeraldoMagela_M.pdf: 6211710 bytes, checksum: 5a88c999846d2e6c89c5a5487e7cbed2 (MD5) Previous issue date: 2013 / Resumo: Agências governamentais ao redor do mundo estão realizando grandes investimentos na utilização de Tecnologia da Informação e Comunicação em suas atividades. Essa tendência, conhecida como Governo eletrônico, impulsiona grande demanda por pesquisas cujo foco principal é o desenvolvimento de aplicações destinadas a um governo mais transparente e colaborativo. Aplicações para esse tipo de cenário introduzem uma série de desafios que precisam ser enfrentados, incluindo maior interoperabilidade entre sistemas, escalabilidade, questões de segurança, entre outros. Nesse sentido, o paradigma de Arquitetura Orientada a Serviços (SOA) apresenta-se como uma interessante proposta para mitigar a heterogeneidade dos serviços prestados pelas diversas entidades envolvidas. Além disso, computação em grade pode ser considerada uma solução promissora para aplicações de middleware em Governo eletrônico, graças à sua alta capacidade de armazenamento e processamento, além de sua recente orientação a serviços, tornando-a uma poderosa ferramenta para aplicações intra-domínio. Considerando esses desafios, este trabalho propõe uma plataforma para aplicações de Governo eletrônico em sistemas em grades utilizando serviços de suporte fornecidos pelo Globus Toolkit 4 (GT4) no contexto da Web Semântica. O trabalho inclui a implementação de um protótipo do middleware e sua validação através de um cenário de aplicação / Abstract: Government agencies around the world are making large investments in the use of Information and Communication Technology in their activities. This trend, known as electronic government, drives a demand for research focused on development of applications aimed at a more transparent and collaborative government. Applications for this type of scenario pose a series of challenges to be faced, including greater interoperability between systems, scalability, and security issues, among others. In this sense, the paradigm of Service-Oriented Architecture (SOA) presents itself as an interesting proposal to mitigate the heterogeneity of services provided by various involved entities. Furthermore, grid computing can be considered a promising solution for middleware applications in e-Government due to its high storage and processing capacity, and also its recent service orientation, making it a powerful tool for intra-domain applications. Considering these challenges, this dissertation proposes a platform for e-Government applications on grid computing, using the support services provided by the Globus Toolkit 4 (GT4) in the context of the Semantic Web. The work includes the implementation of a middleware prototype and its validation through an application scenario / Mestrado / Ciência da Computação / Mestre em Ciência da Computação

Patient’s mobility throughout his lifetime leaves a trial of information scattered in laboratories, clinical institutes, primary care units, and other hospitals. Hence, the medical history of a patient is valuable when subjected to special healthcare units or undergoes home-care/personal-care in elderly stage cases. Despite the rhetoric about patient-centred care, few attempts were made to measure and improve in this arena. In this thesis, we will describe and implement a high-level view of a Patient Centric information management, deploying at a preliminary stage, the use of Agent Technologies and Grid Computing. Thus, developing and proposing an infrastructure that allows us to monitor and survey the patient, from the doctor’s point of view, and investigate a Persona, from the patients’ side, that functions and collaborates among different medical information structures. The Persona will attempt to interconnect all the major agents (human and software), and realize a distributed grid info-structure that directly affect the patient, therefore, revealing an adequate and cost-effective solution for most critical information needs. The results comprehended in the literature survey, consolidating Healthcare Information Management with emerged intelligent Multi-Agent System Technologies (MAS) and Grid Computing; intends to provide a solid basis for further advancements and assessments in this field, by bridging and proposing a framework between the home-care sector and the flexible agent architecture throughout the healthcare domain.

One of the fundamental requirements of Grid computing is efficient and effective resource discovery mechanism. Resource discovery involves discovery of appropriate resources required by user applications. In this regard various resource discovery mechanisms have been proposed during the recent years. These mechanisms range from centralized to hierarchical information servers approach. Most of the techniques developed based on these approaches have scalability and fault tolerance limitations. To overcome these limitations Peer to Peer based discovery mechanisms are proposed. / shairbaz@gmail.com

Les applications Monte-Carlo sont facilement parallélisables, mais une parallélisation efficace sur des grilles de calcul est difficile à réaliser. Des stratégies avancées d'ordonnancement et de parallélisation sont nécessaires pour faire face aux taux d'erreur élevés et à l'hétérogénéité des ressources sur des architectures distribuées. En outre, la fusion des résultats partiels est également une étape critique. Dans ce contexte, l'objectif principal de notre travail est de proposer de nouvelles stratégies pour une exécution plus rapide et plus fiable des applications Monte-Carlo sur des grilles de calcul. Ces stratégies concernent à la fois le phase de calcul et de fusion des applications Monte-Carlo et visent à être utilisées en production. Dans cette thèse, nous introduisons une approche de parallélisation basée sur l'emploi des tâches pilotes et sur un nouvel algorithme de partitionnement dynamique. Les résultats obtenus en production sur l'infrastructure de grille européenne (EGI) en utilisant l'application GATE montrent que l'utilisation des tâches pilotes apporte une forte amélioration par rapport au système d'ordonnancement classique et que l'algorithme de partitionnement dynamique proposé résout le problème d'équilibrage de charge des applications Monte-Carlo sur des systèmes distribués hétérogènes. Puisque toutes les tâches finissent presque simultanément, notre méthode peut être considérée comme optimale à la fois en termes d'utilisation des ressources et de temps nécessaire pour obtenir le résultat final (makespan). Nous proposons également des stratégies de fusion avancées avec plusieurs tâches de fusion. Une stratégie utilisant des sauvegardes intermédiaires de résultat (checkpointing) est utilisée pour permettre la fusion incrémentale à partir des résultats partiels et pour améliorer la fiabilité. Un modèle est proposé pour analyser le comportement de la plateforme complète et aider à régler ses paramètres. Les résultats expérimentaux montrent que le modèle correspond à la réalité avec une erreur relative de 10% maximum, que l'utilisation de plusieurs tâches de fusion parallèles réduit le temps d'exécution total de 40% en moyenne, que la stratégie utilisant des sauvegardes intermédiaires permet la réalisation de très longues simulations sans pénaliser le makespan. Pour évaluer notre équilibrage de charge et les stratégies de fusion, nous mettons en œuvre une simulation de bout-en-bout de la plateforme décrite ci-dessus. La simulation est réalisée en utilisant l'environnement de simulation SimGrid. Les makespan réels et simulés sont cohérents, et les conclusions tirées en production sur l'influence des paramètres tels que la fréquence des sauvegardes intermédiaires et le nombre de tâches de fusion sont également valables en simulation. La simulation ouvre ainsi la porte à des études paramétriques plus approfondies. / Particle-tracking Monte-Carlo applications are easily parallelizable, but efficient parallelization on computing grids is difficult to achieve. Advanced scheduling strategies and parallelization methods are required to cope with failures and resource heterogeneity on distributed architectures. Moreover, the merging of partial simulation results is also a critical step. In this context, the main goal of our work is to propose new strategies for a faster and more reliable execution of Monte-Carlo applications on computing grids. These strategies concern both the computing and merging phases of Monte-Carlo applications and aim at being used in production. In this thesis, we introduce a parallelization approach based on pilots jobs and on a new dynamic partitioning algorithm. Results obtained on the production European Grid Infrastructure (EGI) using the GATE application show that pilot jobs bring strong improvement w.r.t. regular metascheduling and that the proposed dynamic partitioning algorithm solves the load-balancing problem of particle-tracking Monte-Carlo applications executed in parallel on distributed heterogeneous systems. Since all tasks complete almost simultaneously, our method can be considered optimal both in terms of resource usage and makespan. We also propose advanced merging strategies with multiple parallel mergers. Checkpointing is used to enable incremental result merging from partial results and to improve reliability. A model is proposed to analyze the behavior of the complete framework and help tune its parameters. Experimental results show that the model fits the real makespan with a relative error of maximum 10%, that using multiple parallel mergers reduces the makespan by 40% on average, that checkpointing enables the completion of very long simulations and that it can be used without penalizing the makespan. To evaluate our load balancing and merging strategies, we implement an end-to-end SimGrid-based simulation of the previously described framework for Monte-Carlo computations on EGI. Simulated and real makespans are consistent, and conclusions drawn in production about the influence of application parameters such as the checkpointing frequency and the number of mergers are also made in simulation. These results open the door to better and faster experimentation. To illustrate the outcome of the proposed framework, we present some usage statistics and a few examples of results obtained in production. These results show that our experience in production is significant in terms of users and executions, that the dynamic load balancing can be used extensively in production, and that it significantly improves performance regardless of the variable grid conditions.

L’Institut National des Produits Chimiques de l’Académie des Sciences du Vietnam (INPC) développe depuis plusieurs années une activité autour de la recherche de nouveaux médicaments issus de la biodiversité. Le développement d’un nouveau médicament prend de l’ordre d’une dizaine d’années et passe par plusieurs phases. Dans la phase de découverte, l’activité des composés chimiques sur une cible biologique est mesurée afin de mettre en évidence une action inhibitrice. Le développement d’approches in silico pour le criblage virtuel des composés chimiques est une alternative aux approches classiques in vitro beaucoup plus coûteuses à mettre en œuvre. L’utilisation de la grille a été identifiée comme une voie économiquement prometteuse pour accompagner la recherche de nouveaux médicaments au Vietnam. En effet, le développement de nouvelles stratégies basées sur l’utilisation de plates-formes de soumission de tâches (DIRAC, HTCaaS) a permis d’améliorer considérablement le taux de succès et le confort des utilisateurs, ouvrant la voie à une démocratisation de la grille.Dans ce contexte, l’objectif poursuivi dans le cadre de cette thèse est d’étudier dans quelle mesure des plates-formes multidisciplinaires pouvaient répondre aux besoins des chimistes de l’INPC. Le travail s’est concentré sur les modalités d’un partage équitable d’une plate-forme de soumission de tâches sur la grille par une ou plusieurs communautés d’utilisateurs. L’ordonnancement des tâches sur un serveur commun doit permettre que les différents groupes aient une expérience positive et comparable. Sur les infrastructures de grille EGEE et EGI en Europe , on peut distinguer deux grandes catégories d’utilisateurs : les utilisateurs « normaux » qui vont solliciter les ressources pour des tâches requérant typiquement de quelques dizaines à quelques centaines d’heures de calcul, et les « gros » utilisateurs qui vont lancer des grandes productions nécessitant le traitement de plusieurs milliers de tâches pendant des dizaines, voire des centaines de milliers d’heures de calcul. Les stratégies d’ordonnancement déployées aujourd’hui sur les plates-formes comme DIRAC ou HTCaaS ne permettent pas de servir de façon optimale et simultanée ces deux familles d’utilisateurs.Le manuscrit présente une évaluation par simulation des performances de plusieurs stratégies d’ordonnancement des tâches d’une plate-forme soumettant des jobs pilotes. L’outil SimGrid a permis de simuler l’infrastructure de grille régionale déployée en Auvergne à partir de traces archivées de son utilisation. Après évaluation des performances de plusieurs politiques d’ordonnancement tirées de la littérature, une nouvelle politique a été proposée dans laquelle les utilisateurs normaux et les très gros utilisateurs sont gérés de façon indépendante. Grâce à cette politique, le ralentissement expérimenté par les très gros utilisateurs est réduit significativement sans pénaliser excessivement les utilisateurs normaux. L’étude a été étendue à une fédération de clouds utilisant les mêmes ressources et arrive aux mêmes conclusions. Les performances des politiques d’ordonnancement ont ensuite été évaluées sur des environnements de production, à savoir l’infrastructure de grille européenne EGI et l’infrastructure nationale de supercalculateurs de la Corée du Sud. Un serveur DIRAC a été adossé aux ressources de l’organisation virtuelle biomédicale d’EGI pour étudier les ralentissements observés par les utilisateurs de ce serveur. Pareillement, les ralentissements expérimentés par les utilisateurs de la plate-forme HTCaaS au KISTI ont été observés en excellent accord avec les résultats de simulation avec SimGrid.Ces travaux confirment la faisabilité et l’intérêt d’une plate-forme unique au Vietnam au service des communautés scientifiques consommatrices des ressources académiques de grille et de cloud, notamment pour la recherche de nouveaux médicaments. / Virtual Screening (VS) is a computational technique used in the drug discovery process to select the most promising candidate drugs for in vitro testing from millions of chemical compounds. This method can offer an efficient alternative to reduce the cost of drug discovery and platform. The Natural Products Chemistry Institute of the Academy of Sciences of Vietnam (INPC) collects samples from local biodiversity and determines the 3D structure of single molecules. Their challenge is to set up a virtual screening platform on grid computing for their chemists to process their data. However, as the number of users who might have a wide range of virtual screening applications (in terms of the number of tasks and execution time) increases with limited available computing resources, it becomes crucial to devise an effective scheduling policy that can ensure a certain degree of fairness, user satisfaction and overall system throughput. In this context, the thesis focuses on an effective scheduling policy for the virtual screening workflow where multiple users with varying numbers of tasks are actively sharing a common system infrastructure. We have researched in theory and proposed some candidate policies. With the simulation results and the experimentation results in real system, we proposed the best policy for the fairness between users, which can be applied to INPC virtual screening platform.

The concept of grid computing has permeated all areas of distributed computing, changing the way in which distributed systems are designed, developed and implemented. At the same time ‘geobrowsers’, such as Google Earth, NASA World Wind and Virtual Earth, along with in-vehicle navigation, handheld GPS devices and maps on mobile phones, have made interactive maps and geographic information an everyday experience. Behind these maps lies a wealth of spatial data that is collated from a vast number of different sources. A spatial data infrastructure (SDI) aims to make spatial data from multiple sources available to as wide an audience as possible. Current research indicates that, due to a number of reasons, data sharing in these SDIs is still not common. This dissertation presents an analysis of the data grid approach for SDIs. Starting off, two imaginary scenarios spell out for the first time how data grids can be applied to enable the sharing of address data in an SDI. The work in this dissertation spans two disciplines: Computer Science (CS) and Geographic Information Science (GISc). A study of related work reveals that the data grid approach in SDIs is both a novel application for data grids (CS), as well as a novel technology in SDI environments (GISc), and this dissertation advances mutual understanding between the two disciplines. The novel evaluation framework for national address databases in an SDI is used to evaluate existing information federation models against the data grid approach. This evaluation, as well as an analysis of address data in an SDI, confirms that there are similarities between the data grid approach and the requirement for consolidated address data in an SDI. The evaluation further shows that where a large number of organizations are involved, such as for a national address database, and where there is a lack of a single organization tasked with the management of a national address database, the data grid is an attractive alternative to other models. The Compartimos (Spanish for ‘we share’) reference model was developed to identify the components with their capabilities and relationships that are required to grid-enable address data sharing in an SDI. The definition of an address in the broader sense (i.e. not only for postal delivery), the notion of an address as a reference and the definition of an addressing system and its comparison to a spatial reference system contribute towards the understanding of what an address is. A novel address data model shows that it is possible to design a data model for sharing and exchange of address data, despite diverse addressing systems and without impacting on, or interfering with, local laws for address allocation. The analysis in this dissertation confirms the need for standardization of domain specific geographic information, such as address data, and their associated services in order to integrate data from distributed heterogeneous sources. In conclusion, results are presented and recommendations for future work, drawn from the experience on the work in this dissertation, are made. / Thesis (PhD)--University of Pretoria, 2009. / Computer Science / unrestricted

No description available.

This thesis discusses the new way of delivering IT services over the Internet widely known as cloud computing. In its opening part, cloud computing is put into a historical context of the evolution of enterprise computing, and the dominant issues the IT department faces today are mentioned. Further, the paper deals with several components that make up the architecture of cloud computing and reviews the benefits and drawbacks an enterprise can have while it adopts this new model. One of the primary aims of this thesis is to identify the impact of the technology trends on cloud computing. The thesis brings together four major computing trends, namely virtualization, multi-tenant architecture, service-oriented architecture and grid computing. Another aim is to focus on two trends related to IT infrastructure that will lead to fundamental changes in IT industry. The first of them is the emergence of extremely large-scale data centers at low cost locations, which can serve tremendous amount of customers and achieve considerable economies of scale. The second trend this paper points out is the shift from multi-purpose all-in-one computers into a wide range of mobile devices dedicated to a specific user's needs. The last aim of this thesis is to clarify the economic impact of cloud computing in terms of costs and changes in business models. The thesis concludes by evaluating the current adoption and predicting the future trend of cloud computing.