Gorlatch / Danelutto Integrated Research in GRID Computing

CoreGRID Integration Workshop 2005 (Selected Papers) November 28-30, Pisa, Italy
1. Auflage 2007
ISBN: 978-0-387-47658-2
Verlag: Springer
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 284 Seiten

The aim of CoreGRID is to strengthen and advance scientific and technological excellence in the area of Grid and Peer-to-Peer technologies in order to overcome the current fragmentation and duplication of effort in this area. To achieve this objective, the workshop brought together a critical mass of well-established researchers from a number of institutions which have all constructed an ambitious joint program of activities. Priority in the workshop was given to work conducted in collaboration between partners from different research institutions and to promising research proposals that could foster such collaboration in the future.

Autoren/Hrsg.

Gorlatch, Sergei

Danelutto, Marco

Inhaltsverzeichnis

1;Contents;6
2;Foreword;8
3;Contributing Authors;12
4;DATA INTEGRATION AND QUERY REFORMULATION IN SERVICE- BASED GRIDS;20
4.1;1. Introduction;21
4.2;2. Data Integration in Grids;22
4.3;3. XMAP: A Decentralized XML Data Integration Framework;23
4.4;4. Introduction to Grid query processing services;26
4.5;5. Integrating the XMAP algorithm in service-based Grids: A walk-through example;27
4.6;6. XPath to OQL mapping;29
4.7;7. Implementation Roadmap: Service Interactions and System Design;30
4.8;8. Summary;31
4.9;Acknowledgments;31
4.10;References;31
5;TOWARDS A COMMON DEPLOYMENT MODEL FOR GRID SYSTEMS;34
5.1;1. Introduction;35
5.2;2. ASSIST and GridCCM Software Component Models;36
5.2.1;2.1 Assist;36
5.2.2;2.2 GridCCM: a Parallel Component Model;38
5.2.3;2.3 Discussion;40
5.3;3. General Overview of the Deployment Process;40
5.3.1;3.1 Application Submission;41
5.3.2;3.2 Resource Discovery;42
5.3.3;3.3 Resource Selection;42
5.3.4;3.4 Deployment Planning;43
5.3.5;3.5 Deployment Enactment;43
5.3.6;3.6 Application Execution;44
5.4;4. Current Prototypes;44
5.4.1;4.1 GEA;44
5.4.2;4.2 Adage;46
5.4.3;4.3 Comparison of GEA and Adage;47
5.5;5. Conclusion;47
5.6;References;48
6;TOWARDS AUTOMATIC CREATION OF WEB SERVICES FOR GRID COMPONENT COMPOSITION;50
6.1;1. Introduction;51
6.2;2. Higher-Order Components (HOCs);52
6.3;3. Higher-Order Components (HOCs) built upon ProActive/Fractal;55
6.4;4. Accessing HOC components via ProActive Web services;57
6.5;5. Conclusion and Perspectives;59
6.6;Acknowledgments;60
6.7;References;60
7;ADAPTABLE PARALLEL COMPONENTS FOR GRID PROGRAMMING;62
7.1;1. Introduction;63
7.2;2. Components and Adaptation;64
7.2.1;2.1 Higher-Order Components (HOCs);65
7.2.2;2.2 Example: The Farm-HOC;65
7.2.3;2.3 The Implementation of Adaptable HOCs;66
7.3;3. Case Study: Sequence Alignment;67
7.4;4. Adaptations with Globus & WSRF;68
7.4.1;4.1 Enabling Mobile Code;69
7.4.2;4.2 Customizing the Farm-HOC for Sequence Alignment;70
7.4.3;4.3 Adapting the Farm-HOC to the Wavefront Pattern;71
7.5;5. Experimental Results;73
7.6;6. Conclusion and Related Work;74
7.7;Acknowledgments;75
7.8;References;76
8;SKELETON PARALLEL PROGRAMMING AND PARALLEL OBJECTS;78
8.1;1. Introduction;79
8.2;2. Parallel Object-Oriented Programming;79
8.3;3. Structured parallel programming with ASSIST;81
8.3.1;3.1 Grid Application Deployment;82
8.4;4. Objects and skeletons getting along;83
8.4.1;4.1 Same memory for ASSIST and POP-C++;84
8.4.2;4.2 ASSIST components written in POP-C++;84
8.4.3;4.3 Deploying ASSIST and POP-C++ alike;85
8.5;5. Architecture for a common deployer;86
8.6;6. Conclusion;88
8.7;References;89
9;TOWARDS THE AUTOMATIC MAPPING OF ASSIST APPLICATIONS FOR THE GRID;92
9.1;1. Introduction;93
9.2;2. The ASSIST environment and its run-time support;94
9.2.1;2.1 The ASSIST coordination language;95
9.2.2;2.2 The ASSIST run-time support;95
9.2.3;2.3 Towards fully grid-aware applications;96
9.3;3. Introduction to performance evaluation and PEPA;96
9.4;4. Performance models of ASSIST applications;98
9.4.1;4.1 The ASSIST application;98
9.4.2;4.2 The PEPA model;98
9.4.3;4.3 Automatic generation of the model;99
9.4.4;4.4 Performance results;101
9.4.5;4.5 Analysis summary;103
9.4.6;4.6 Future work;104
9.5;5. Conclusions;104
9.6;Acknowledgments;105
9.7;References;105
10;AN ABSTRACT SCHEMA MODELING ADAPTIVITY MANAGEMENT;108
10.1;1. An Abstract Schema for Adaptation;109
10.2;2. Adaptivity;110
10.3;3. Example of the abstract decomposition;112
10.4;4. Dynaco/AFPAC: a generic framework for developers to manage adaptation;113
10.4.1;4.1 Dynaco: generic dynamic adaptation framework;113
10.4.2;4.2 AFPAC: dynamic adaptation of parallel components;114
10.5;5. ASSIST: Managing dynamicity using language and compilation approaches;115
10.6;6. A comparative discussion;119
10.7;7. Conclusions;120
10.8;Acknowledgments;120
10.9;References;120
11;A FEEDBACK- BASED APPROACH TO REDUCE DUPLICATE MESSAGES IN UNSTRUCTURED PEER- TO- PEER NETWORKS;122
11.1;1. Introduction;123
11.2;2. Related work;124
11.3;3. The Feedback-based algorithm;125
11.4;4. Random vs. small-world graphs;129
11.5;5. Experimental results on static graphs;130
11.6;6. Experimental results on dynamic graphs;134
11.7;7. Conclusions;136
11.8;Acknowledgments;136
11.9;References;136
12;FAULT- INJECTION AND DEPENDABILITY BENCHMARKING FOR GRID COMPUTING MIDDLEWARE;138
12.1;1. Introduction;139
12.2;2. Related Works;139
12.2.1;2.1 Fault-injection;139
12.2.2;2.2 Dependability benchmarking;140
12.3;3. Our proposal;141
12.3.1;3.1 FAIL-FCI;141
12.3.2;3.2 QUAKE: A Dependability Benchmark Tool for Grid Services;143
12.4;4. Experimental Results;145
12.4.1;4.1 Fault Injection;145
12.4.2;4.2 Dependability Benchmarking;146
12.5;5. Conclusions and Current Status;151
12.6;Acknowledgments;151
12.7;References;151
13;USER MANAGEMENT FOR VIRTUAL ORGANIZATIONS;154
13.1;1. Introduction;155
13.2;2. Definitions;155
13.3;3. Existing Solutions;156
13.3.1;3.1 Perun;156
13.3.2;3.2 Virtual User System;157
13.3.3;3.3 VOMS, LCAS and LCMAPS;157
13.3.4;3.4 Virtual Workspaces, Runtime Environments, Dynamic Virtual Environments;157
13.4;4. System Requirements;158
13.4.1;4.1 Authentication;158
13.4.2;4.2 Authorization;158
13.4.3;4.3 Encapsulation of Jobs and Results;159
13.4.4;4.4 Accounting and Logging Facilities;159
13.4.5;4.5 Other Requirements;160
13.5;5. Proposed Solution;161
13.5.1;5.1 Virtual Environment Management Service;161
13.5.2;5.2 Virtual Environment Database;162
13.6;6. Summary;164
13.7;7. Acknowledgment;164
13.8;References;164
14;ON THE INTEGRATION OF PASSIVE AND ACTIVE NETWORK MONITORING IN GRID SYSTEMS;166
14.1;1. Introduction;167
14.2;2. Classification of Network Monitoring Techniques;168
14.2.1;2.1 Link versus Path Monitoring;168
14.2.2;2.2 Passive versus Active Monitoring;169
14.3;3. Passive Network Monitoring for Grid Infrastructures;170
14.3.1;3.1 Metrics based on a Single Observation Point;171
14.3.2;3.2 Metrics based on Multiple Observation Points;173
14.4;4. Active Network Monitoring for Grid Infrastructures;174
14.5;5. The Domain Overlay Database;175
14.5.1;5.1 Monitoring Activities Description;176
14.6;6. Security and Privacy Issues;178
14.7;7. Summary and Conclusions;179
14.8;References;180
15;SENSOR ORIENTED GRID MONITORING INFRASTRUCTURES FOR ADAPTIVE MULTI- CRITERIA RESOURCE MANAGEMENT STRATEGIES;182
15.1;1. Introduction;183
15.2;2. Motivation;184
15.3;3. Related Works and Activities;184
15.4;4. Embedding Sensors in Applications - MPI Example;186
15.5;5. Event and Altert Monitoring;187
15.6;6. Example Adaptive Multi-criteria Resource Management Strategic;189
15.7;7. Preliminary Results and Future Work;190
15.8;Acknowledgments;192
15.9;References;192
16;TOWARDS SEMANTICS- BASED RESOURCE DISCOVERY FOR THE GRID*;194
16.1;1. Introduction;195
16.2;2. Background;196
16.2.1;2.1 Semantic Description of Grid Services;196
16.2.2;2.2 Matching Services;197
16.3;3. Architecture;199
16.4;4. Implementation;201
16.5;5. Evaluation;202
16.6;6. Conclusions and Future Work;204
16.7;Acknowledgments;205
16.8;References;205
17;SCHEDULING WORKFLOWS WITH BUDGET CONSTRAINTS^;208
17.1;1. Introduction;209
17.2;2. Background;210
17.3;3. The Algorithm;212
17.3.1;3.1 Outline;212
17.3.2;3.2 The LOSS Approach;213
17.3.3;3.3 The GAIN Approach;214
17.3.4;3.4 Variants;215
17.4;4. Experimental Results;215
17.4.1;4.1 Experiment Setup;215
17.4.2;4.2 Results;216
17.5;5. Conclusion;220
17.6;References;221
18;INTEGRATION OF ISS INTO THE VIOLA METASCHEDULING ENVIRONMENT;222
18.1;1. Introduction;223
18.2;2. UNICORE and the Meta-scheduling Service;223
18.2.1;2.1 UNICORE;224
18.2.2;2.2 Meta-Scheduling Service;224
18.3;3. Intelligent Scheduling System Model;225
18.3.1;3.1 Application types;225
18.3.2;3.2 The I' model;226
18.4;4. Resulting Grid IMiddleware Architecture;226
18.4.1;4.1 Meta-Scheduling Service;226
18.4.2;4.2 Resource Broker;227
18.4.3;4.3 Data Warehouse;227
18.4.4;4.4 System Information;228
18.4.5;4.5 Monitoring Module;228
18.5;5. Detailed Scheduling Scenario;228
18.6;6. Application Example: Submission of ORBS;230
18.7;7. Conclusion;232
18.8;Acknowledgments;232
18.9;References;232
19;MULTI- CRITERIA GRID RESOURCE MANAGEMENT USING PERFORMANCE PREDICTION TECHNIQUES;234
19.1;1. Introduction;235
19.2;2. Related work;236
19.3;3. Workload;236
19.4;4. Prediction System;237
19.4.1;4.1 Architecture;237
19.4.2;4.2 Method;238
19.5;5. Multi-criteria prediction-based resource selection;239
19.6;6. Preliminary Results;241
19.7;7. Conclusion;241
19.8;Acknowledgments;243
19.9;References;243
20;A PROPOSAL FOR A GENERIC GRID SCHEDULING ARCHITECTURE*;246
20.1;1. Introduction;247
20.2;2. Grid Scheduling Scenarios;248
20.2.1;2.1 Scenario I: Enterprise Grids;248
20.2.2;2.2 Scenario II: High Performance Computing Grids;248
20.2.3;2.3 Scenario III: Global Grids;249
20.3;3. Common functions of Grid Scheduling;249
20.4;4. Scheduling Instance;253
20.5;5. Conclusion;257
20.6;References;257
21;GRID SUPERSCALAR ENABLED P- GRADE PORTAL;260
21.1;1. Introduction;261
21.2;2. P-GRADE Portal;262
21.3;3. GRID superscalar;264
21.3.1;3.1 GRID superscalar monitor;267
21.4;4. Comparison of P-GRADE Portal and GRID superscalar;268
21.5;5. Overview of the solution;269
21.6;6. Conclusions, related and future work;272
21.7;Acknowledgments;272
21.8;References;272
22;REDESIGNING THE SEGL PROBLEM SOLVING ENVIRONMENT: A CASE STUDY OF USING MEDIATOR COMPONENTS;274
22.1;1. Introduction;275
22.2;2. Component-based Grid application environments;275
22.3;3. The SEGL system architecture;279
22.4;4. Extracting mediator components from the SEGL functionality;282
22.5;5. Related Work and Ongoing Developments;284
22.5.1;5.1 Related Work;284
22.5.2;5.2 Ongoing Developments of Mediator Components;285
22.6;6. Conclusions;286
22.7;Acknowledgements;286
22.8;References;286
23;SYNTHETIC GRID WORKLOADS WITH IBIS, KOALA, AND GRENCHMARK;290
23.1;1. Introduction;291
23.2;2. A Case for Synthetic Grid Workloads;291
23.2.1;2.1 Analytical IModeling and Simulations;291
23.2.2;2.2 Experimental Testing;292
23.2.3;2.3 Grid Applications Types;293
23.2.4;2.4 Purposes of Synthetic Grid Workloads;293
23.3;3. An Extensible Framework for Grid Synthetic Workloads;294
23.3.1;3.1 Ibis: Grid Applications;295
23.3.2;3.2 GRENCHMARK: Synthetic Grid Workloads;295
23.3.3;3.3 Using the Framework;296
23.4;4. A Concrete Case: Synthetic Workloads for the DAS;297
23.4.1;4.1 KOALA: ScheduUng Grid AppUcations;297
23.4.2;4.2 The Workload Generation;297
23.4.3;4.3 The Workload Submission;298
23.5;5. The Experimental Results;298
23.5.1;5.1 The Performance Results;299
23.5.2;5.2 Dealing With Errors;299
23.6;6. Proposed Research Roadmap;300
23.7;7. Conclusions and Ongoing Work;300
23.8;Acknowledgments;301
23.9;References;301
24;Author Index;303

Leseproben

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.