Integrated Research in GRID Computing (eBook)

Contents 6
Foreword 8
Contributing Authors 12
DATA INTEGRATION AND QUERY REFORMULATION IN SERVICE- BASED GRIDS 20
1. Introduction 21
2. Data Integration in Grids 22
3. XMAP: A Decentralized XML Data Integration Framework 23
4. Introduction to Grid query processing services 26
5. Integrating the XMAP algorithm in service-based Grids: A walk-through example 27
6. XPath to OQL mapping 29
7. Implementation Roadmap: Service Interactions and System Design 30
8. Summary 31
Acknowledgments 31
References 31
TOWARDS A COMMON DEPLOYMENT MODEL FOR GRID SYSTEMS 34
1. Introduction 35
2. ASSIST and GridCCM Software Component Models 36
2.1 Assist 36
2.2 GridCCM: a Parallel Component Model 38
2.3 Discussion 40
3. General Overview of the Deployment Process 40
3.1 Application Submission 41
3.2 Resource Discovery 42
3.3 Resource Selection 42
3.4 Deployment Planning 43
3.5 Deployment Enactment 43
3.6 Application Execution 44
4. Current Prototypes 44
4.1 GEA 44
4.2 Adage 46
4.3 Comparison of GEA and Adage 47
5. Conclusion 47
References 48
TOWARDS AUTOMATIC CREATION OF WEB SERVICES FOR GRID COMPONENT COMPOSITION 50
1. Introduction 51
2. Higher-Order Components (HOCs) 52
3. Higher-Order Components (HOCs) built upon ProActive/Fractal 55
4. Accessing HOC components via ProActive Web services 57
5. Conclusion and Perspectives 59
Acknowledgments 60
References 60
ADAPTABLE PARALLEL COMPONENTS FOR GRID PROGRAMMING 62
1. Introduction 63
2. Components and Adaptation 64
2.1 Higher-Order Components (HOCs) 65
2.2 Example: The Farm-HOC 65
2.3 The Implementation of Adaptable HOCs 66
3. Case Study: Sequence Alignment 67
4. Adaptations with Globus & WSRF
4.1 Enabling Mobile Code 69
4.2 Customizing the Farm-HOC for Sequence Alignment 70
4.3 Adapting the Farm-HOC to the Wavefront Pattern 71
5. Experimental Results 73
6. Conclusion and Related Work 74
Acknowledgments 75
References 76
SKELETON PARALLEL PROGRAMMING AND PARALLEL OBJECTS 78
1. Introduction 79
2. Parallel Object-Oriented Programming 79
3. Structured parallel programming with ASSIST 81
3.1 Grid Application Deployment 82
4. Objects and skeletons getting along 83
4.1 Same memory for ASSIST and POP-C++ 84
4.2 ASSIST components written in POP-C++ 84
4.3 Deploying ASSIST and POP-C++ alike 85
5. Architecture for a common deployer 86
6. Conclusion 88
References 89
TOWARDS THE AUTOMATIC MAPPING OF ASSIST APPLICATIONS FOR THE GRID 92
1. Introduction 93
2. The ASSIST environment and its run-time support 94
2.1 The ASSIST coordination language 95
2.2 The ASSIST run-time support 95
2.3 Towards fully grid-aware applications 96
3. Introduction to performance evaluation and PEPA 96
4. Performance models of ASSIST applications 98
4.1 The ASSIST application 98
4.2 The PEPA model 98
4.3 Automatic generation of the model 99
4.4 Performance results 101
4.5 Analysis summary 103
4.6 Future work 104
5. Conclusions 104
Acknowledgments 105
References 105
AN ABSTRACT SCHEMA MODELING ADAPTIVITY MANAGEMENT 108
1. An Abstract Schema for Adaptation 109
2. Adaptivity 110
3. Example of the abstract decomposition 112
4. Dynaco/AFPAC: a generic framework for developers to manage adaptation 113
4.1 Dynaco: generic dynamic adaptation framework 113
4.2 AFPAC: dynamic adaptation of parallel components 114
5. ASSIST: Managing dynamicity using language and compilation approaches 115
6. A comparative discussion 119
7. Conclusions 120
Acknowledgments 120
References 120
A FEEDBACK- BASED APPROACH TO REDUCE DUPLICATE MESSAGES IN UNSTRUCTURED PEER- TO- PEER NETWORKS 122
1. Introduction 123
2. Related work 124
3. The Feedback-based algorithm 125
4. Random vs. small-world graphs 129
5. Experimental results on static graphs 130
6. Experimental results on dynamic graphs 134
7. Conclusions 136
Acknowledgments 136
References 136
FAULT- INJECTION AND DEPENDABILITY BENCHMARKING FOR GRID COMPUTING MIDDLEWARE 138
1. Introduction 139
2. Related Works 139
2.1 Fault-injection 139
2.2 Dependability benchmarking 140
3. Our proposal 141
3.1 FAIL-FCI 141
3.2 QUAKE: A Dependability Benchmark Tool for Grid Services 143
4. Experimental Results 145
4.1 Fault Injection 145
4.2 Dependability Benchmarking 146
5. Conclusions and Current Status 151
Acknowledgments 151
References 151
USER MANAGEMENT FOR VIRTUAL ORGANIZATIONS 154
1. Introduction 155
2. Definitions 155
3. Existing Solutions 156
3.1 Perun 156
3.2 Virtual User System 157
3.3 VOMS, LCAS and LCMAPS 157
3.4 Virtual Workspaces, Runtime Environments, Dynamic Virtual Environments 157
4. System Requirements 158
4.1 Authentication 158
4.2 Authorization 158
4.3 Encapsulation of Jobs and Results 159
4.4 Accounting and Logging Facilities 159
4.5 Other Requirements 160
5. Proposed Solution 161
5.1 Virtual Environment Management Service 161
5.2 Virtual Environment Database 162
6. Summary 164
7. Acknowledgment 164
References 164
ON THE INTEGRATION OF PASSIVE AND ACTIVE NETWORK MONITORING IN GRID SYSTEMS 166
1. Introduction 167
2. Classification of Network Monitoring Techniques 168
2.1 Link versus Path Monitoring 168
2.2 Passive versus Active Monitoring 169
3. Passive Network Monitoring for Grid Infrastructures 170
3.1 Metrics based on a Single Observation Point 171
3.2 Metrics based on Multiple Observation Points 173
4. Active Network Monitoring for Grid Infrastructures 174
5. The Domain Overlay Database 175
5.1 Monitoring Activities Description 176
6. Security and Privacy Issues 178
7. Summary and Conclusions 179
References 180
SENSOR ORIENTED GRID MONITORING INFRASTRUCTURES FOR ADAPTIVE MULTI- CRITERIA RESOURCE MANAGEMENT STRATEGIES 182
1. Introduction 183
2. Motivation 184
3. Related Works and Activities 184
4. Embedding Sensors in Applications - MPI Example 186
5. Event and Altert Monitoring 187
6. Example Adaptive Multi-criteria Resource Management Strategic 189
7. Preliminary Results and Future Work 190
Acknowledgments 192
References 192
TOWARDS SEMANTICS- BASED RESOURCE DISCOVERY FOR THE GRID* 194
1. Introduction 195
2. Background 196
2.1 Semantic Description of Grid Services 196
2.2 Matching Services 197
3. Architecture 199
4. Implementation 201
5. Evaluation 202
6. Conclusions and Future Work 204
Acknowledgments 205
References 205
SCHEDULING WORKFLOWS WITH BUDGET CONSTRAINTS^ 208
1. Introduction 209
2. Background 210
3. The Algorithm 212
3.1 Outline 212
3.2 The LOSS Approach 213
3.3 The GAIN Approach 214
3.4 Variants 215
4. Experimental Results 215
4.1 Experiment Setup 215
4.2 Results 216
5. Conclusion 220
References 221
INTEGRATION OF ISS INTO THE VIOLA METASCHEDULING ENVIRONMENT 222
1. Introduction 223
2. UNICORE and the Meta-scheduling Service 223
2.1 UNICORE 224
2.2 Meta-Scheduling Service 224
3. Intelligent Scheduling System Model 225
3.1 Application types 225
3.2 The I' model 226
4. Resulting Grid IMiddleware Architecture 226
4.1 Meta-Scheduling Service 226
4.2 Resource Broker 227
4.3 Data Warehouse 227
4.4 System Information 228
4.5 Monitoring Module 228
5. Detailed Scheduling Scenario 228
6. Application Example: Submission of ORBS 230
7. Conclusion 232
Acknowledgments 232
References 232
MULTI- CRITERIA GRID RESOURCE MANAGEMENT USING PERFORMANCE PREDICTION TECHNIQUES 234
1. Introduction 235
2. Related work 236
3. Workload 236
4. Prediction System 237
4.1 Architecture 237
4.2 Method 238
5. Multi-criteria prediction-based resource selection 239
6. Preliminary Results 241
7. Conclusion 241
Acknowledgments 243
References 243
A PROPOSAL FOR A GENERIC GRID SCHEDULING ARCHITECTURE* 246
1. Introduction 247
2. Grid Scheduling Scenarios 248
2.1 Scenario I: Enterprise Grids 248
2.2 Scenario II: High Performance Computing Grids 248
2.3 Scenario III: Global Grids 249
3. Common functions of Grid Scheduling 249
4. Scheduling Instance 253
5. Conclusion 257
References 257
GRID SUPERSCALAR ENABLED P- GRADE PORTAL 260
1. Introduction 261
2. P-GRADE Portal 262
3. GRID superscalar 264
3.1 GRID superscalar monitor 267
4. Comparison of P-GRADE Portal and GRID superscalar 268
5. Overview of the solution 269
6. Conclusions, related and future work 272
Acknowledgments 272
References 272
REDESIGNING THE SEGL PROBLEM SOLVING ENVIRONMENT: A CASE STUDY OF USING MEDIATOR COMPONENTS 274
1. Introduction 275
2. Component-based Grid application environments 275
3. The SEGL system architecture 279
4. Extracting mediator components from the SEGL functionality 282
5. Related Work and Ongoing Developments 284
5.1 Related Work 284
5.2 Ongoing Developments of Mediator Components 285
6. Conclusions 286
Acknowledgements 286
References 286
SYNTHETIC GRID WORKLOADS WITH IBIS, KOALA, AND GRENCHMARK 290
1. Introduction 291
2. A Case for Synthetic Grid Workloads 291
2.1 Analytical IModeling and Simulations 291
2.2 Experimental Testing 292
2.3 Grid Applications Types 293
2.4 Purposes of Synthetic Grid Workloads 293
3. An Extensible Framework for Grid Synthetic Workloads 294
3.1 Ibis: Grid Applications 295
3.2 GRENCHMARK: Synthetic Grid Workloads 295
3.3 Using the Framework 296
4. A Concrete Case: Synthetic Workloads for the DAS 297
4.1 KOALA: ScheduUng Grid AppUcations 297
4.2 The Workload Generation 297
4.3 The Workload Submission 298
5. The Experimental Results 298
5.1 The Performance Results 299
5.2 Dealing With Errors 299
6. Proposed Research Roadmap 300
7. Conclusions and Ongoing Work 300
Acknowledgments 301
References 301
Author Index 303

1. Introduction (p. 32)
The Grid vision introduced in the end of the nineties has now become a reality with the availability of quite a few Grid infrastructures, most of them experimental but some others will come soon in production. Although most of the research and development efforts have been spent in the design of Grid middleware systems, the question of how to program such large scale computing infrastructures remains open. Programming such computing infrastructures will be quite complex considering its parallel and distributed nature.

The programmer vision of a Grid infrastructure is often determined by its programming model. The level of abstraction that is proposed today is rather low, giving the vision either of a parallel machine, with a message-passing layer such as MPI, or a distributed system with a set of services, such as Web Services, to be orchestrated. Both approaches offer a very low level programming abstraction and are not really adequate, limiting the spectrum of applications that could take benefit from Grid infrastructures.

Of course such approaches may be sufficient for simple applications but a Grid infrastructure has to be generic enough to also handle complex applications with ease. To overcome this situation, it is required to propose high level abstractions to facilitate the programming of Grid infrastructures and in a longer term to be able to develop more secure and robust next generation Grid middleware systems by using these high level abstractions for their design as well. The current situation is very similar to what happened with computers in the sixties: minimalist operating systems were developed first with assembly languages before being developed, in the seventies, by languages that offer higher levels of abstraction.

Several research groups are already investigating how to design or adapt programming models that provide this required level of abstraction. Among these models, component-oriented programming models are good candidates to deal with the complexity of programming Grid infrastructures. A Grid application can be seen as a collection of components interconnected in a certain way that must be deployed on available computing resources managed by the Grid infrastructure.

Components can be reused for new Grid applications, reducing the time to build new applications. However, from our experience such models have to be combined with other programming models that are required within a Grid infrastructure. It is imaginable that a parallel program can be encapsulated within a component. Such a parallel program is based on a parallel programming model which might be for instance message-based or skeletonbased.

Moreover, a component oriented programming model can be coupled with a service oriented approach exposing some component ports as services through the use of Web Services. The results of this is that this combination of several models to design Grid applications leads to a major challenge: the deployment of applications within a Grid infrastructure.

Such programming models are always implemented through various runtime or middleware systems that have their own dependencies vis-a-vis of operating systems, making it extremely challenging to deploy applications within a heterogeneous environment, which is an intrinsic property of a Grid infrastructure.

The objective of this paper is to propose a common deployment process based on the experience gained from the ASSIST and GridCCM projects.