Palestras Técnicas / Technical Presentations
|KEYNOTE: - MyProxy: A Multi-Purpose Grid
Jim Basney (NCSA)
Abstract. MyProxy is a Globus Toolkit component for managing certificates and private keys that that has been used by the grid community since 2000 in projects such as EGEE, EU DataGrid, Earth System Grid, FusionGrid, LHC Computing Grid, NASA Information Power Grid, NEESgrid, Open Science Grid, and TeraGrid. In the past year, we have added the ability for MyProxy to act as an online Certificate Authority (CA) in addition to a credential repository and have added support for additional authentication methods via PAM and the Pubcookie web single-sign on system. These new capabilities allow MyProxy to support new grid authentication scenarios. In this talk, I will describe these new MyProxy capabilities and how they are being used in grid projects.
Jim Basney is a Senior Research Scientist at the National Center for Supercomputing Applications at the University of Illinois, where hecleads the MyProxy project. His research focuses on the security,creliability, efficiency, and usability of distributed systems.cDr. Basney obtained his Doctorate and Masters degrees in ComputercSciences from the University of Wisconsin-Madison.
|Invited Talk: E-Infrastructure shared between
Europe and Latin America – EELA
Diego Carvalho (UFRJ)
Abstract. The EELA project is building a bridge between consolidated e-Infrastructure initiatives in Europe and emerging ones in Latin America by creating a Human Collaboration Network in e-Science oriented to the dissemination of knowledge in Grid Computing. Furthermore, a pilot Grid infrastructure is being deployed in LA integrated into the European one, as an element of know-how transfer and to support the proof-of-concept of enhanced scientific applications of interest for both research communities. The EELA Consortium is formed by institutions from Spain, Portugal, Italy, Argentina, Brazil, Chile, Cuba, Venezuela, Peru and Mexico.
|Uma proposta para execução
distribuída de consultas em um ambiente de Grid para o CoDIMS
Gustavo Gaburro Trevisol, Alvaro Cesar Pereira Barbosa (UFES)
Abstract. Grid is a computational environment in which applications can use multiple distributed computational resources in a safe, coordinated, efficient and transparent way. Data integration systems are distributed and they can make use of Grid environments to obtain a better performance and a rational use of available resources. This work describes a proposal for the Distributed Query Execution Engine (MECD) to CoDIMS. The MECD is inserted in a Grid Environment for distributed and parallel execution of sub-queries sent to data sources. This proposal will permit a more efficient execution of queries that requires more processing power.
|Desafios para Provisão de Integridade de
Processamento em Grades Computacionais
Felipe Martins, Márcio Maia, Rossana de C Andrade, Aldri dos Santos, J Neuman de Souza (UFC)
Abstract. Highly distributed and dynamic environments such as computational grids are constantly under attacks, for example, malicious attacks that corrupt job results. This paper focuses on the main challenges for guaranteeing the integrity of processing in grids, presenting a classification of misbehaving nodes and well-known fault tolerance techniques that can be applied to detect and minimize the presence of malicious nodes. It also compares the most important grid simulators that could be extended to evaluate the behavior of such hostile environment. Furthermore, a case study shows the efficiency of the investigated techniques.
|Escalonamento Tolerante a Falhas na
Recuperação de Aplicações em MPI
Idalmis Milián Sardiña, Cristina Bôeres, Lúcia Drummond (UFF)
Abstract. A crucial problem in distributed systems is the probability to the occurrence of failures in resources. Recent studies search different forms to improve application execution time, also including fault-tolerant mechanisms, but in many cases testing its politicies on simulated environments. This work presents an MPI tool to execute parallel applications on a real architecture, recovering the application execution using fault-tolerant scheduling techniques. For such, it considers the information generated by a static scheduling heuristic, offering mechanisms for automatic failure detection and efficient application recovery.
|ZeliGrid: uma arquitetura para a
implantação de aplicações com requisitos
não-funcionais dinâmicos em grades computacionais
Rodrigo Souza Granja, Alexandre Sztajnberg (UERJ)
Abstract. The management of resources is an important aspect to be considered regarding applications that might have different non-functional or operational requirements, when running in distributed and heterogeneous environments, such as the Computational Grids. In this context it is necessary to provide means to specify the required resource constraints and an infrastructure that can adapt the applications in face of changes on the recource avalilability. In this paper we adopt the CR-RIO approach to deploy parallel applications that have non-functional requirements. To provide the supporting infrastructure we integrate some of the Globus services and NWS monitoring services to the concept of QoS Contracts in the form of an architecture. Deployment and implementation details are discussed.
|MAG, um middleware de grade baseado em agentes: estado
atual e perspectivas futuras
Francisco da Silva e Silva, Rafael Lopes, Bysmarck de Sousa, Antônio Viana, Stanley de Sousa (UFMA)
Abstract. MAG (Mobile Agents for Grid Computing Environments) explores the mobile agent paradigm as a way to overcome the construction challenges of computational grids. MAG executes Grid applications by dynamically loading the application code into mobile agents. The MAG agent can be dynamically reallocated to Grid nodes through a transparent migration mechanism, as a way to provide load balancing and support for non-dedicated nodes. MAG includes mechanisms for fault tolerance, pervasive grid, and data grid. We make extensive use of the agent paradigm to design and implement MAG components, forming a multi agent infrastructure for computational Grids.
|Sistema de Monitoramento/Gerência de Recursos e de
Segurança para Grades Computacionais Baseadas no GlobusToolKit
Luis de O Gonçalves, Fábio Fagundes, Bruno Schulze, Letícia de Sá, Thais de Mello (LNCC)
Abstract. This document presents a monitoring tool for the resources and security of Grids based on the Globus Toolkit. Different from other solutions, it avoids the installation of additional software, since all the collection process is carried out by scripts, which are sent using the communication channel provided by the Globus libraries. Scripts will also be used to provide the necessary mechanisms for the security adjustments.
|NextComp - Molecular Dynamics Application for Long-Range
Interacting Systems on a Computational Grid Environment
Marcelo de Albuquerque, Alexandre de Almeida, Leonardo Lessa, Márcio de Albuquerque, Nilton Alves, Luis G Moyano (CBPF), Constantino Tsallis (CBPF e Santa Fe Institute)
Abstract. This paper presents the ongoing activities of the NextComp project. This project has as main objective to investigate the validity of nonextensive statistical mechanics for systems with long-range interactions, in particular the Hamiltonian Mean Field model. This will be achieved through molecular dynamics (MD) simulations by a parallel program taking advantages from Charm++ language. Development and tests are carried out in a Linux cluster environment, being our interest focused on a grid environment in order to overcome the limiting factors of the simulations.
|Sand Castle: infra-estrutura segura para hospedar
dinamicamente aplicações em grades computacionais
Leonardo Mattes, João Antonio Zuffo (USP)
Abstract. The grid paradigm aims a better exploitation of the computational resources in a distributed and flexible systems, however its utilization brings new challenges regarding security. One of the foreseen grid functionalities creates dynamic jobs and services, as well as operation bringings flexibility to the system that should be exploited for installations of “malicious” applications. The present work presents Sand Castle, a tool that aims to host and control Java applications in grid computing environments, and, once it is hosted, its applications has to be monitored and controlled in according to security polices, that has been established by the grid environment.
|Integração de Ambientes Heterogêneos
de Grades Computacionais
Alessandra Nascimento, Marcio Miranda (UFG)
Abstract. Grid computing provides integration of different environments, creating a unified system, where organizations might share, manage and access different resources regardless where they are localized. However, the lack of integration among different grid computing systems does not provide total exploitation of resources. This work investigates and develops an interface for integration of two heterogeneous grid computing environments: Globus and InteGrade. Such integration increases the computational power, transparently. Users will be able to submit jobs through both grid environments and applications will be executed into Globus, or into InteGrade or into both systems.
|Modelo para Integração de Sistemas de
Detecção de Intrusão através de Grids
Paulo Fernando da Silva, Carlos Becker Westphall, Carla Merkle Westphall (UFSC)
Resumo. Este artigo apresenta um modelo para integração de IDSs usando Grids computacionais. O modelo proposto possibilita que IDSs heterogêneos trabalhem em conjunto formando um DIDS. São realizadas considerações e simulações sobre o modelo proposto demonstrando seu funcionamento.
|ch_hyperbone: Um Dispositivo para Execução
de Programas MPI em Hipercubos Virtuais
Samuel Mello, Rafael Caiuta (UFPR), Luis Bona (UTFPR), Elias Duarte Jr.(UFPR), Keiko Fonseca (UTFPR)
Abstract. Several applications rely on distributed and parallel computing as a powerful tool for solving problems that require large amounts of resources. HyperBone is an overlay network based on virtual hypercubes that abstracts the heterogeneity of connected resources, providing message routing and delivery, and fault management. This paper presents the implementation of a software device that allows the execution of parallel programs based on the MPI standard on top of the HyperBone platform. Experimental results show the functionality of the device, the impact caused by the use of the hypercube topology on the throughput and evaluates the performance of an image rendering application.
|ContextGrid - Uma infra-estrutura de suporte a contexto
para integração de dispositivos móveis a grades
Alaor José da Silva Junior, Márcio Nunes de Miranda, Fábio Moreira Costa (UFG)
Resumo. Este trabalho descreve uma arquitetura de computação sensível ao contexto, que provê a base para a integração de dispositivos móveis em ambientes de computação em grade, permitindo a adaptação dinâmica do middleware de acordo com as variações de contexto a que estes dispositivos estão sujeitos.
|Um Ambiente Hierárquico para
Definição e Negociação de Recursos Ociosos
em uma Grade Institucional
Lourival Aparecido de Góis (UTFPR), Walter da Cunha Borelli (UNICAMP)
Resumo. Este artigo apresenta um modelo hierárquico de busca e disponibilização de recursos ociosos, localizados em uma Instituição de
ensino com vários campi, distribuídos e interconectados através de um backbone confiável e gerenciado por regras de QoS.
|Mobilidade em Ambientes de Grades Computacionais
Hans Alberto Franke, Carlos Becker Westphall, Carlos Oberdan Rolim, Fabio Navarro, Fernando Koch (UFSC)
Resumo. Este artigo visa o estudo de um ambiente hospitalar que utiliza grades computacionais, para coleta e disponibilização de dados vitais dos pacientes em tempo real. Propõe-se a criação de um middleware que permita a integração dos dados coletados através dos sensores com a grade computacional, permitindo assim uma homogeneização dos nodos da grade e uma melhoria quanto à facilidade oferecida ao usuário. Este middleware deve possuir os mínimos recursos necessários existentes em outros middlewares, suportando também dispositivos móveis.
|Gerência e Fornecimento de Certificados de
Confiança para Nós Móveis Usando Certificados de
Fabio L. Licht (LNCC e IME-RJ), Bruno Schulze (LNCC), Edison Ishikawa (IME-RJ)
Abstract.This work presents a proposal to extend an existing service for short lived certificates (SLCs), with purpose of allowing mobile nodes to enter computational grids. Amongst the parts of this work, it is cited, the use public, private keys, supply of trustworthy certificates and dynamic SLCs. MyProxy is the tool responsible for supplying dynamic certificates and greater agility in the inclusion of new users and also the extension to nodes that want to enter a computational grid.
|Análise Paralela e Distribuída de Dados
Micrometeorológicos Utilizando a Plataforma JXTA
Marcelo Veiga Neves (UFRGS), Tiago Scheid, Andrea Schwertner Charão, Guilherme Sausen Welter, Osvaldo Luiz Leal de Moraes (UFSM)
|Specification of a MyProxy Plugin for Mozilla
Luiz Manoel Rocha Gadelha Júnior (LNCC)
Abstract. This poster specifies a MyProxy plugin for the Mozilla web browser. It should allow grid users to generate and store their credentials in the cryptographic repository of the browser and to delegate proxy credentials to MyProxy repositories.