Table of contents : Title page Preface Organization Table of Contents CoreGRID Workshop onGRID Middleware Introduction Architecture of a Network Monitoring Element Introduction Inside View of a Network Monitoring Element The Passive Network Monitoring Component Outline of a Secure Group Membership Scheme Related Work Conclusions Support for Automatic Diagnosis and DynamicConfiguration of Scalable Storage Systems Introduction Global Reconfiguration Directed Reconfiguration Related Work Conclusions Adding Dynamism to OGSA-DQP:Incorporating the DynaSOAr Framework inDistributed Query Processing Introduction Brief Description of OGSA-DQP DynaSOAr Architecture Dynamic OGSA-DQP Implementation Experiment Setup Results and Analysis Related Work Current and Future Directions Conclusion Review of Security Models Applied toDistributed Data Access Introduction Definitions Grid Data Management Systems Security Requirements Data Types Data Management Security Technologies Authentication Authorization Local Security Enforcement Building a Grid Data Access Framework Conclusions Security Requirements Analysis for Large-ScaleDistributed File Systems Introduction FileStamp Architecture Security Requirements of FileStamp Generic Requirements Specific Requirements Solutions for the FileStamp Security Requirements Authentication Authorization Availability, Confidentiality, and Integrity Resilience and Fault Tolerance Data Lifecycle Management (DLM) Conclusions References Coupling Contracts forDeployment on Alien Grids Introduction Related Work Coupling Matches with Contracts Clause Types Typed Clauses Coupling Interfaces Coupling Contracts Matching Parties: Descriptors and Applications Example Coupling Distributed Applications with Deployment on the Grid Background on ProActive Deployment Descriptors The Problematic of Applications and Descriptors Clause Types Clauses in ProActive Descriptors Clauses in ProActive Applications Constraints Conclusions and Future Work A Transparent Framework for HierarchicalMaster-Slave Grid Computing Introduction Related Work TMS Framework Reflection-Based Implementation MOP-Based Implementation Programming Model User QoS Requirements Framework Evaluation Conclusion References A Multi-level Scheduler for the Grid ComputingYML Framework Introduction YML Framework YML Advantages Component Creation Application Creation with YvetteML A Multi-level Scheduling Model in YML An Economic Model Scheduling Scenario Access Policy Case Study Features for the Scheduling Model Conclusions and Perspectives Virtual Environments - Framework forVirtualized Resource Access in the Grid Introduction Comparison of VA and VM Virtual Workspaces Approach Architecture of the Framework Accounting and Audit Conclusions Grid Meta-Broker Architecture:Towards an Interoperable GridResource Brokering Service Introduction Related Work Abstract Architecture Conclusions A Super-Peer Model for Multiple JobSubmission on a Grid Introduction Job Assignment and Data Download Performance Evaluation Conclusions References A Scheduling Algorithm for High PerformancePeer-to-Peer Platform Introduction What Is a Real Peer-to-Peer System? XtremWeb The Coordinator Workers XtremWeb-CH Automatic Execution of Parallel and Distributed Applications Granularity and Scheduling Direct Communication Monitoring Tools Experimental Measures Conclusion References Brokering Multi-grid Workflowsin the P-GRADE Portal Introduction The P-GRADE Portal Workflow Management in the P-GRADE Portal Workflow Notation Developing and Editing Workflows Workflow Execution Workflow Management in Details Multi-grid Workflow Brokering Related Work Summary and Conclusions Diet: New Developments and Recent Results Introduction DIET Architecture DIET Scheduling Plug-In Schedulers Collectors of Resource Information DIET Batch Scheduler Management DIET Workflow Management DIET Deployment DIET Fault-Tolerance Fault Detection MA Topology Recovery Checkpoint/Restart Mechanism DIET Visualization and Large Scale Validation DIET Applications A BLAST Application Using DIET Cosmological Simulation with RAMSES and Galics Conclusion and Future Work Execution Support of High PerformanceHeterogeneous Component-Based Applicationson the Grid Introduction Component Deployment in a Multi-middleware Heterogenous Environment Related Work The ASSIST Environment Architecture Grid Execution Agent Design Core Deployment Cycle A Modular Multi-middleware Architecture Conclusion and Future Work Towards a Grid Information Knowledge Base Introduction A Core Grid Ontology Framework Building a Grid Knowledge Base Using CGO Querying a Grid Knowledge Base Conclusions UNICORE Summit 2006 Introduction A Versatile Execution Management System forNext-Generation UNICORE Grids Introduction Experiences with UNICORE The UNICORE NJS: A Gap Analysis Requirements for a Next-Generation NJS The XNJS: Design and Implementation of a Next-Generation UNICORE NJS Core Architecture Execution Management Security Status of the XNJS Conclusions and Outlook Towards More Flexible and Increased Securityand Privacy in Grids General Remarks Identification and Authorisation in an Organisation Identification Attributes of Entities Authorisation Organisations in the Grid Identification and Authentication on the Grid Authorisation Based on Identity Authorisations Based on Properties of the Entities User Database at the Grid Resource Site Managing Authorisation for VOs The Attributes of Requestors in Their Organisation The Attributes of Requestors in the VO Privacy --- Anonymity Attributes of the Resources Authorisation Consequences for the UNICORE Development Authentication A First Use of Attributes: The Explicit Trust Delegation The Proposed Authorisation Architecture Attribute Authorities Authorisation Authorisation in the Non-UNICORE World VOMS Shibboleth GridShib PERMIS GridShibPERMIS Conclusion Integration of Grid Cost Model into ISS/VIOLAMeta-scheduler Environment Introduction Application Component Characteristics Embarrassingly Parallel Application Components Application Components with Point-to-Point Communications Application Components with Multicast Communication Needs Components Demanding Shared Memory Meta-scheduling Features The Broker Action List Decision: Grid Cost Model Epilogue: Prepare Data for Next Execution Simulator to Tune Parameters Conclusions A One-Stop, Fire-and-(Almost)Forget,Dropping-Off and Rendezvous Point Introduction Atomic Services and Interoperation Functionality Java Business Integration Orchestration Domain Specific Languages BPEL Rules Summary Grid-Based Processing of High-VolumeMeteorological Data Sets Introduction Challenges to Be Addressed Transfer of Large Data Sets Parallelization Solution Approaches Employing Grid Technologies Globus Toolkit 4 Application Scenario. Data Transfers with Globus Toolkit 4. Condor Application Scenario. The Condor Approach. UNICORE Conclusions and Future Work BLAST Application on the GPE/UnicoreGSGrid Introduction BLAST Software Grid Programming Environment (GPE) BLAST GridBean for GPE Conclusions and Future Work Job Management Enterprise Application Introduction Requirements for a Job Management Application Architecture and Implementation of the Job Management Enterprise Application (JMEA) The Arcon Client Library The Components Security Conclusion and Outlook UNICORE Deployment Within the DEISASupercomputing Grid Infrastructure Introduction DEISA Infrastructure Overview UNICORE UNICORE Components UNICORE and DEISA Infrastructure Integration UNICORE Deployment for DEISA Integration of UNICORE with DEISA Batch Scheduling Systems and GPFSMulti Cluster The Final Picture Conclusions References Petascale ComputationalBiology and Bioinformatics Introduction Progress in Scaling Biomolecular Simulations toPetaflop Scale Platforms Introduction Experiences with Blue Gene/L Peta-scale Challenges Molecular Simulation Validity Performance and Scalability Algorithmic Explorations Conclusions Progress Towards Petascale Applications inBiology: Status in 2006 Introduction Methods and Materials Results Discussion Conclusion Toward a Solution of the Reverse EngineeringProblem Using FPGAs Introduction Dealing with Large Genetic Networks Finite Field Multiplication A New FPGA-Based Approach Experimental Results Conclusions and Future Work Two Challenges in Genomics That Can Benefitfrom Petascale Platforms Introduction Genomic Signatures for Monitoring the Rate of Accumulation of Somatic Mutations Optimal Combinations of Genomic Signatures for Human HLA Typing Conclusions High Throughput Image Analysis onPetaFLOPS Systems High Throughput Image Analysis Performance and Limitations Suggested Changes to Scale Up to PetaFLOPS Systems Introduce Distributed Storage of Data Improve Communication and Job Control Author Index