The GENI Book (eBook)

Rick McGeer, Mark Berman, Chip Elliott, Robert Ricci (Herausgeber)

eBook Download: PDF

2016 | 1st ed. 2016
XXX, 651 Seiten
Springer International Publishing (Verlag)
978-3-319-33769-2 (ISBN)

Infos

This book, edited by four of the leaders of the National Science Foundation's Global Environment and Network Innovations (GENI) project, gives the reader a tour of the history, architecture, future, and applications of GENI.  Built over the past decade by hundreds of leading computer scientists and engineers, GENI is a nationwide network used daily by thousands of computer scientists to explore the next Cloud and Internet and the applications and services they enable, which will transform our communities and our lives. Since by design it runs on existing computing and networking equipment and over the standard commodity Internet, it is poised for explosive growth and transformational impact over the next five years.

Over 70 of the builders of GENI have contributed to present its development, architecture, and implementation, both as a standalone US project and as a federated peer with similar projects worldwide, forming the core of a worldwide network. Applications and services enabled by GENI, from smarter cities to intensive collaboration to immersive education, are discussed.  The book also explores the concepts and technologies that transform the Internet from a shared transport network to a collection of 'slices' -- private, on-the-fly application-specific nationwide networks with guarantees of privacy and responsiveness.

The reader will learn the motivation for building GENI and the experience of its precursor infrastructures, the architecture and implementation of the GENI infrastructure, its deployment across the United States and worldwide, the new network applications and services enabled by and running on the GENI infrastructure, and its international collaborations and extensions. This book is useful for academics in the networking and distributed systems areas, Chief Information Officers in the academic, private, and government sectors, and network and information architects.

Rick McGeer is a Principal Investigator with the Communications Design Group at SAP Labs, Chief Scientist at US Ignite, and an Adjunct Professor of Computer Science at the University of Victoria.  He earned a PhD in Computer Science from the University of California, Berkeley, in 1989.  He is a Senior Member of the IEEE and a Member of the ACM.  He is the author of over 100 papers in the fields of networking, distributed systems, programming language design, formal verification, timing analysis and logic design.  He is the author of Integrating Functional and Temporal Analysis of Logic Circuits.

Mark Berman is Vice President for Technology Development at Raytheon BBN Technologies and GENI Project Director. Mark works with the GENI community, which spans dozens of universities, government and industry partners, to ensure that GENI is well designed, technically feasible, and satisfies its research requirements. Mark's research interests are in the area of complex distributed systems and their usability. He has twice (1966, 2006) been named Time's Person of the Year (shared).

Chip Elliott is Chief Scientist at Raytheon BBN Technologies, an Adjunct Professor of Computer Science at Dartmouth College, and a Fellow of the AAAS, ACM, and IEEE. He served as the first GENI Project Director.

Robert Ricci is a Research Assistant Professor in the School of Computing at the University of Utah, and one of the directors of the Flux Research Group. He earned a PhD from the University of Utah in 2010, and an Honors BS from Utah in 2001. His research interests are in the fields of systems and networking, and much of his energy has gone into creating top-quality experimental environments. He has worked in a diverse set of areas including distributed systems, combinatorial optimization, security, networking, simulation, and embedded systems. He has been a primary architect and implementor of Emulab and follow-on systems such as ProtoGENI and CloudLab since 2000.

Inhalt

Introduction 8
Background: Why GENI? 8
How Did GENI Come To Be? 12
GENI's Community Development Approach 15
Organization of the Book 18
References 19
Acknowledgements 22
Contents 24
Contributors 28
Part I Precursors 32
The GENI Vision: Origins, Early History, Possible Futures 34
1 The Original Idea of GENI 34
1.1 The Objective 34
1.2 Expansion of the Objective 35
1.3 Origins of the GENI Idea 36
1.4 Motivations 38
1.5 Overview of GENI's Early Days (Years) 39
2 What Has Shaped the GENI Project? 41
3 What Does the Future Hold? 44
3.1 Funding 45
3.2 Community Involvement 45
3.3 Leadership 46
3.4 Industry Engagement 47
3.5 Useful Experimentation 47
4 Conclusion 48
Precursors: Emulab 49
1 Running Experiments on Emulab 52
2 The Emulab Control Infrastructure 53
3 Distinguishing Features of Emulab 55
3.1 Focus on Scientific Fidelity 55
3.2 Focus on Multi-Tenant, Bare-Metal Allocation 56
3.3 The Network as a First-Class Entity 57
4 The Evolution of Emulab into ProtoGENI 58
5 Lessons from Emulab 59
6 The Future of Emulab 60
References 61
DETERLab and the DETER Project 64
1 Introduction 64
2 Project History 65
2.1 Project Evolution 65
3 Objectives 66
4 DETERLab Technologies 69
4.1 Core Technologies 69
4.2 Containers for Scale and Fidelity 70
4.3 Federation 73
4.4 Experiment Orchestration 75
4.5 Multi Party Experiments 79
4.6 Modeling Human Behavior 81
4.6.1 The Dash Agent Platform 82
5 A DETERLab Use Case 83
6 DETERLab in Education 86
7 Looking to the Future 87
8 Conclusion 89
References 89
ORBIT: Wireless Experimentation 92
1 Introduction 92
2 Design Requirements 95
3 ORBIT Testbed Technical Details 98
3.1 ORBIT System 98
3.2 ORBIT Hardware 100
3.3 ORBIT Software 107
3.4 ORBIT Experiment Life-Cycle 111
4 Experimental Research Enabled by ORBIT 115
4.1 Radio Channel Signature Based Encryption 116
4.2 Dynamic Spectrum Coordination in Dense Multi-Radio Environments 117
4.3 Global Name Resolution Service (GNRS) for Future Internet 119
5 ORBIT Evolution and Future Upgrades 120
6 Links to GENI Project 122
References 123
Part II Architecture and Implementation 125
GENI Architecture Foundation 128
1 Introduction 128
1.1 Facilitating Trusted Exchange of Resources 128
2 GENI Federation 130
3 Trust Foundation 131
4 GENI Concepts 132
5 GENI Services 133
5.1 Federation API 133
5.2 Aggregate Manager (AM) API 136
5.2.1 GENI Resource Specifications (RSpecs) 138
5.3 Monitoring Services 139
6 Tools 139
6.1 The GENI Portal 140
7 Summary 141
References 143
The Need for Flexible Community Research Infrastructure 144
1 Introduction 144
2 Meta-Infrastructure 146
3 Risk and Cost Reduction 149
4 Maximizing Research 150
5 Conclusion 152
References 152
A Retrospective on ORCA: Open Resource Control Architecture 154
1 Introduction 154
2 Overview of the ORCA Platform 156
2.1 Resource Leases 157
2.2 Extension Modules 158
2.3 Leasing Engine 159
2.4 Resource Descriptions 160
2.5 Building Aggregates with ORCA 162
3 Orchestration and Cross-Aggregate Resource Control 163
3.1 ORCA Resource Control Plane 164
3.2 Brokers 165
3.3 Controllers 166
3.4 Automated Stitching and Topology Mapping 167
3.5 GENI Proxy Controller 168
4 Reflections on GENI and ORCA 169
4.1 Platforms vs. Products + Protocols 169
4.2 Federation 170
4.3 Orchestration 171
4.4 Description Languages 172
References 173
Programmable, Controllable Networks 175
1 Integrating the Network into IT Infrastructure 175
2 The OpenFlow Protocol 180
2.1 Brief Summary of OpenFlow and the OpenFlow Protocol 181
2.2 Promises of OpenFlow 183
3 Initial Implementations and Campus Experiments 186
4 Using OpenFlow in a Multi-Tenant Network 188
4.1 Hybrid Switching 189
4.2 FlowVisor 190
4.3 Software Datapaths 191
5 Integration with GENI 192
6 Experimenter Experience 193
6.1 Fundamental Infrastructure Issues 193
6.2 Virtualization/Slicing Issues 194
6.3 Lessons for the Future 195
7 New Opportunities with OpenFlow and SDN 196
8 SDN: The Next Generation 199
References 201
4G Cellular Systems in GENI 205
1 Introduction 205
2 Deployment 207
2.1 Spiral III Deployment Sites 208
2.2 Spiral IV Deployment Sites 208
3 Typical Deployment Architecture 209
3.1 NEC WiMAX Base Station 209
3.2 Airspan WiMAX Base Station (Fig. 4) 211
3.3 LTE Base Stations 212
3.4 4G Client Devices 213
4 GENI Wireless Site Management Framework 215
4.1 RF Aggregate Manager 217
4.2 Datapath Management 217
4.3 Virtualization 218
4.4 Monitoring 220
4.5 Portal Integration and Account Federation 221
4.6 Integration with GENI Rack 221
5 Experimentation 222
6 Extending GENI Cellular Coverage Using SciWiNet 225
References 226
Authorization and Access Control: ABAC 228
1 Introduction 228
2 GENI Authorization Requirements 232
2.1 GENI Authorization Needs 232
3 Attribute Based Access Control and ABAC 233
3.1 ABAC and GENI 235
3.2 ABAC Logics 235
3.3 RT0 Logic 236
3.4 RT1 and RT2 238
3.5 Case Study: GENI Authorization and Speaks-for 240
3.5.1 Semantics of GENI Privilege Credentials 241
3.5.2 GENI Policy in RT1 242
4 Implementing ABAC—The libabac System 243
4.1 System Design Issues 244
4.1.1 Principal Requirements 244
4.1.2 Representing Attributes and Rules 246
4.1.3 Negotiation 247
4.1.4 Negotiating with Sensitive Data 248
4.2 libabac Software System Architecture 249
4.2.1 Basic Operation 249
4.2.2 Asynchronous Public Data 251
4.2.3 Controlling Sensitive Credentials 252
4.2.4 Representing Partial Proofs 252
4.3 Integration 253
4.4 A libabac Implementation 253
4.4.1 Core Objects 255
4.4.2 Interfaces for Creating, Managing, and Utilizing Identities and Credentials 255
4.5 libabac Adoption 256
4.5.1 Use in GENI 257
4.5.2 Use in DETER 257
5 Conclusions and Future Directions 258
References 258
The GENI Experiment Engine 260
1 Introduction and Motivation 260
2 A User's View of the GEE 262
3 GEE Architecture 265
4 GEE Implementation 268
4.1 The GEE Portal 268
4.2 The GEE Compute Service 270
4.3 Fabric and Ansible: Single Pane-of-Glass Control and Configuration 273
4.4 The GEE File System 273
4.5 The GEE Message Service 274
4.6 The GEE Reverse Proxy Service 275
4.7 The GEE Network 276
5 Deploying an Application on GEE 277
6 Related Work 277
7 Conclusions 279
References 280
Part III The GENI National Buildout 282
The GENI Mesoscale Network 284
1 Introduction 284
2 Early Design Activities 285
3 Nationwide Layer-2 Dataplane Network in GENI 288
3.1 Internet2 292
3.2 NLR 295
4 Regional and International Networks in GENI 297
5 Campus Networks in GENI 299
6 VLAN Stitching 300
Reference 302
ExoGENI: A Multi-Domain Infrastructure-as-a-Service Testbed 303
1 Introduction 303
2 Overview: A Testbed of Federated IaaS Providers 305
2.1 Operational Principles 305
2.2 Hardware and Topology 307
2.3 ORCA ExoGENI Control Software 309
2.4 Deployed Software Ecosystem 313
3 ExoGENI Services 315
3.1 Compute Resources 316
3.2 Storage Resources 320
3.3 Rack-Local VLANs and OpenFlow 321
3.4 Transit Network Providers 322
3.5 Network Exchange Points and Multi-Point Connections 324
3.6 Stitchports 326
4 ExoGENI Administration 326
4.1 Scalable Configuration 327
4.2 Monitoring 328
4.3 Administration 330
5 ExoGENI User Tools 330
5.1 Flukes 330
5.2 Supporting Compatibility with GENI Tools 332
6 ExoGENI and the GENI Federation 334
6.1 Relationship to Other GENI Testbeds 334
6.2 Aggregates 335
6.3 GENI Federation: Coordinators 335
6.4 Integration with GENI 336
7 Conclusion 336
References 337
The InstaGENI Project 340
1 Introduction and Motivation 340
2 InstaGENI's Place in the Universe 344
3 Architecture of InstaGENI 346
3.1 The InstaGENI Software Architecture 346
4 The InstaGENI Network 350
5 Implementation of InstaGENI 353
6 Deployment of InstaGENI 356
7 Operations and Maintenance 356
8 Experience and Status 358
9 Related Work 360
10 Conclusions, Extensions and Further Work 361
References 363
Part IV GENI Experiments and Applications 367
The Experimenter's View of GENI 370
1 Useful GENI Concepts 372
1.1 GENI Resources and Resource Aggregates 372
1.2 GENI RSpecs and the GENI AM API 374
1.3 Slice 375
1.4 GENI Projects 376
2 The GENI Experimenter Workflow 377
2.1 Design and Setup Experiment 377
2.2 Execute Experiment 379
2.3 Finish Experiment 380
3 Case Study: GENI Cinema, Implementing an Advanced Service on GENI 381
3.1 Designing GENI Cinema 381
3.2 Use of Software Defined Networking 382
3.3 Deploying GENI Cinema 384
3.4 Connecting Users to GENI Cinema 385
4 Experimenter Tools 386
4.1 RSpec Creation Tools 386
4.1.1 Jacks and Flack 387
4.1.2 jFed 387
4.1.3 geni-lib 387
4.2 Resource Reservation Tools 388
4.2.1 Omni 389
4.2.2 The GENI Portal 389
4.3 Experiment Orchestration and Scripting Tools 390
4.3.1 OEDL 391
4.4 Instrumentation and Measurement Tools 391
4.4.1 GENI Desktop 391
4.4.2 LabWiki 391
4.5 Software Installation and Resource Configuration 392
4.5.1 Install and Execute Scripts 393
4.5.2 Custom Images 393
4.5.3 Configuration Management Tools 394
4.6 Archiving 394
4.6.1 The GENI iRODS Service 394
5 Experiment Repeatability and Reproducibility 394
5.1 Making Experiments Repeatable and Reproducible 395
5.1.1 Reducing Variability Across Runs of an Experiment 395
5.1.2 Sharing Experiment Artifacts for Reprodicibility 395
6 Scaling Up Experiments 396
7 Collaboration 397
7.1 Mechanisms for Collaboration 397
References 397
The GENI Desktop 401
1 Running Experiments in GENI 401
1.1 The Need for Higher-Level Tools and Services 402
1.2 History of the GENI Desktop 403
2 GENI Desktop Design Goals 404
3 GENI Desktop 405
3.1 An Example Workflow 406
3.2 Designing and Creating an Experimental Network (Slice) 407
3.3 Creating Superslices 408
3.4 Running and Interacting with an Experiment 409
3.5 Monitoring an Experiment 411
3.6 Tearing Down an Experiment and Archiving the Results 412
3.7 GENI Desktop Command Line Interface 413
3.8 Components of the GENI Desktop System 414
4 Using the GENI Desktop 415
4.1 The GENI Desktop GUI 415
4.2 The GENI Desktop CLI 421
4.3 Common Usage Models 422
5 Interacting with Other Tools and Services 423
6 Summary 425
References 425
A Walk Through the GENI Experiment Cycle 427
1 Introduction 427
2 LabWiki Overview 429
3 LabWiki User Experience 430
4 Experiment Overview 432
5 Experimental Facilities 432
6 Experiment Design 434
7 Experiment Description and Instrumentation 435
7.1 Describing an Experiment 436
7.2 Instrumenting Resources 438
7.3 The Prepare Panel 439
8 Resource Selection and Provisioning 440
8.1 Process Overview 440
8.2 Labwiki Topology Editor Plugin 441
8.3 The SliceService 443
9 Running an Experiment Trial 443
9.1 The Execute Panel 444
9.2 The JobService and Its Scheduler 444
9.3 Orchestrating Resources 445
9.4 Collecting Measurements 445
10 Result Analysis Over Multiple Trials 446
11 Store and Publish 446
11.1 Storing and Sharing with LabWiki 447
11.2 Publish as a Practical Lab for a Course 448
12 Conclusion 449
References 450
GENI in the Classroom 452
1 Instructor Resources 454
1.1 GENI Course Modules 454
1.2 GENI Wireless Classroom as a Service: Testbed-Hosted Lab Exercises to Challenge Students' Assumptions About Computer Networks 458
1.2.1 Lab Modules 461
2 GENI MOOC: Expanding Access to Lab Experiments in Computer Networks 462
2.1 Design Considerations 462
3 GENI in K-12 STEM Education 465
3.1 The Mars Game 465
3.2 Remote Interactive Digital Cinema Microscope 466
References 468
The Ignite Distributed Collaborative Scientific Visualization System 469
1 Sending Programs to Data and People 470
2 Collaborative Visualization Systems and Distributed Clouds 472
3 Architecture of the IDCVS 476
3.1 The GENI Experiment Engine 477
3.2 The Lively Web 478
3.3 The IDCVS Messaging System 479
3.4 Deployment 481
4 The Atmospheric Quality Visualization System 481
4.1 The Data Server 484
5 Quantitative Analysis of the Problem Space 484
6 Related Work 486
7 Demo and Future Work 487
7.1 Demonstration at Future Internet Summit 487
7.2 Current Operational Deployment 489
7.3 Further Work on the IDCVS 490
7.4 Further Work on The Atmospheric Quality Visualization Application 491
References 492
US Ignite and Smarter Communities 496
1 Genesis of the US Ignite Initiative 496
2 What's New about the Next Generation of the Internet 498
3 The Magic of (the) GENI 511
4 The New Questions 513
5 A Design for a Smart City Metropolitan Internet 513
6 Example Applications and Services 515
7 Connected Collaboration 515
8 Digital Cinema Microscope 516
9 Virtual Reality Educational Tools in Even Low-Income Homes 517
10 Reducing the Cost of 3D Printing 518
11 Collaborative Pollution Viewer 519
12 A Vision for Smarter Cities 519
13 Additional Photos 521
Part V GENI and the World 528
Europe's Mission in Next-Generation Networking with Special Emphasis on the German-Lab Project 530
1 Introduction 530
2 The European Arena 532
2.1 Introducing the European FP7 ICT Approach 533
2.2 The Objective 1.1 535
2.2.1 4Ward 536
2.2.2 Trilogy 537
2.2.3 Publish-Subscribe Internet Routing Paradigm (PSIRP) 538
2.2.4 Network of the Future (Euro-NF) 539
2.3 The FIRE Projects 541
2.3.1 Inside FIRE 542
2.4 Future-Internet Assembly (FIA) 550
3 The Public Private Partnership (PPP) Approach 550
4 The German-Lab Project 551
4.1 The G-Lab Experimental Facility 554
4.2 The Topology Management Tool 556
5 Conclusion and Outlook 558
References 560
SAVI Testbed for Applications on Software-Defined Infrastructure 562
1 Future Applications Marketplace 562
2 Applications on Multitier Clouds with a Smart Edge 564
3 Smart Edge Based on Software-Defined Infrastructure 566
4 The Janus Management System for SDI 569
5 The SAVI Canadian Testbed 573
6 Applications and Experiments on the SAVI Testbed 575
7 Concluding Remarks 577
References 578
Research and Development on Network Virtualization Technologies in Japan: VNode and FLARE Projects 580
1 Introduction 580
2 Brief History of Network-Virtualization Research 582
3 Supporting Deep Programmability 585
4 Problem Definition 587
5 Overview of Network Virtualization Platform 587
6 Research and Development: Strategy, Target, and Development Items 588
6.1 Network-Virtualization Platform Architecture Ensuring Evolution 589
6.2 Control and Management Mechanism: Abstraction and Elasticity of Resources 591
6.3 Resource Control: Securing Resource Independence/Isolation 593
6.4 Improvement in Programmability 595
6.4.1 Compatibility of Programmability with Performance 595
6.4.2 Programmability Extension by Applying Plug-in Functional Modules 597
6.4.3 Edge-Virtualization Technologies: Programming Technologies for Network Access 598
6.4.4 Gateway Function Improvement: Authentication Capability 599
7 Deployment on Testbed and International Federation 601
8 Concluding Remark and Perspectives for the Future 602
References 604
Creating a Worldwide Network for the Global Environment for Network Innovations (GENI) and Related Experimental Environments 606
1 Introduction 607
2 Overview of Chapter 608
3 Required Services 609
4 Global Environment for Network Innovations (GENI) and Related Initiatives 611
5 Basic Architectural Considerations 612
6 Creating a Common International Network Language and Network Programming Languages 615
7 Existing International Facilities 616
7.1 Global Lambda Integrated Facility (GLIF) 616
8 Network Service Interface—NSI Connection Service 617
9 The International GENI (iGENI) and the International Advanced Network Research Facility 619
10 Research Activities and Experiments Conducted Among Current International Environments 621
10.1 Slice Around the World Initiative 621
11 International V-Node 624
12 ToMaTo a Virtual Research Environment for Large Scale Distributed Systems Research 625
13 Monitoring OpenFlow Slices with Ethernet OAM 627
14 Multipath TCP (MPTCP) 627
15 Provider Backbone Bridging Based Network Virtualization 628
16 The Sea Cloud Innovation Environment 629
17 International Multi-Domain Automatic Network Topology Discovery (MDANTD) 632
18 Future Internet Virtualization Environment and VLAN Transit 634
19 Interdomain ExoGENI 638
20 Content Routing 638
21 Brazilian Future Internet Experimental Environment 639
22 High Performance Digital Media Network (HPDMnet) 640
23 Anticipated Future Services and Resource Expansions 641
24 Software Defined Networking Exchanges (SDXs) 642
25 Software Defined Infrastructure (SDI) and Cloud Testbed Integration 644
26 Emerging Architecture and Design Trends for Anticipated Future Facilities 645
27 Conclusions 645
References 646
Appendix: Additional Readings 650
Afterword: A Fire in the Dark 668

Produkt-Details

• Erscheint lt. Verlag: 31.8.2016
• Zusatzinfo: XXX, 651 p. 255 illus., 225 illus. in color.
• Verlagsort: Cham
• Sprache: englisch
• Themenwelt: Informatik ► Weitere Themen ► Hardware
• Schlagworte: Deep programmability • Distributed cloud • FIDC testbeds • GENI • Infrastructure-as-a-service • Network stitching • Network testbeds • OpenFlow • Programmable Networks • Research clouds • slice • Software-defined infrastructure • software-defined networking • Virtual infrastructure
• ISBN-10: 3-319-33769-6 / 3319337696
• ISBN-13: 978-3-319-33769-2 / 9783319337692