分布式系统 原理与范例PDF电子书下载

1 INTRODUCTION 1

1.1 DEFINITION OF A DISTRIBUTED SYSTEM 2

1.2 GOALS 4

1.2.1 Connecting Users and Resources 4

1.2.2 Transparency 5

1.2.3 Openness 8

1.2.4 Scalability 10

1.3 HARDWARE CONCEPTS 16

1.3.1 Multiprocessors 17

CONTENTS 17

PREFACE 17

1.3.2 Homogeneous Multicomputer Systems 19

1.3.3 Heterogeneous Multicomputer Systems 21

1.4 SOFTWARE CONCEPTS 22

1.4.1 Distributed Operating Systems 22

1.4.2 Network Operating Systems 33

1.4.3 Middleware 36

1.5 THE CLIENT-SERVER MODEL 42

1.5.1 Clients and Servers 42

1.5.2 Application Layering 46

1.5.3 Client-Server Architectures 50

1.6 SUMMARY 53

2 COMMUNICATION 57

2.1 LAYERED PROTOCOLS 58

2.1.1 Lower-Level Protocols 61

2.1.2 Transport Protocols 63

2.1.3 Higher-Level Protocols 66

2.2 REMOTE PROCEDURE CALL 68

2.2.1 Basic RPC Operation 69

2.2.2 Parameter Passing 73

2.2.3 Extended RPC Models 77

2.2.4 Example:DCE RPC 80

2.3 REMOTE OBJECT INVOCATION 85

2.3.1 Distributed Objects 86

2.3.2 Binding a Client to an Object 88

2.3.3 Static versus Dynamic Remote Method Invocations 90

2.3.4 Parameter Passing 91

2.3.5 Example 1:DCE Remote Objects 93

2.3.6 Example 2:Java RMI 95

2.4.1 Persistence and Synchronicity in Communication 99

2.4 MESSAGE-ORIENTED COMMUNICATION 99

2.4.2 Message-Oriented Transient Communication 104

2.4.3 Message-Oriented Persistent Communication 108

2.4.4 Example:IBM MQSeries 115

2.5 STREAM-ORIENTED COMMUNICATION 119

2.5.1 Support for Continuous Media 120

2.5.2 Streams and Quality of Service 123

2.5.3 Stream Synchronization 127

2.6 SUMMARY 130

3 PROCESSES 135

3.1.1 Introduction to Threads 136

3.1 THREADS 136

3.1.2 Threads in Distributed Systems 141

3.2 CLIENTS 145

3.2.1 User Interfaces 145

3.2.2 Client-Side Software for Distribution Transparency 147

3.3 SERVERS 149

3.3.1 General Design Issues 149

3.3.2 Object Servers 152

3.4.1 Approaches to Code Migration 158

3.4 CODE MIGRATION 158

3.4.2 Migration and Local Resources 163

3.4.3 Migration in Heterogeneous Systems 165

3.4.4 Example:D'Agents 168

3.5 SOFTWARE AGENTS 173

3.5.1 Software Agents in Distributed Systems 173

3.5.2 Agent Technology 175

3.6 SUMMARY 178

4 NAMING 183

4.1.1 Names,Identifiers,and Addresses 184

4.1 NAMING ENTITIES 184

4.1.2 Name Resolution 189

4.1.3 The Implementation of a Name Space 194

4.1.4 Example:The Domain Name System 201

4.1.5 Example:X.500 206

4.2 LOCATING MOBILE ENTITIES 210

4.2.1 Naming versus Locating Entities 210

4.2.2 Simple Solutions 212

4.2.3 Home-Based Approaches 216

4.2.4 Hierarchical Approaches 217

4.3 REMOVING UNREFERENCED ENTITIES 225

4.3.1 The Problem of Unreferenced Objects 225

4.3.2 Reference Counting 227

4.3.3 Reference Listing 231

4.3.4 Identifying Unreachable Entities 232

4.4 SUMMARY 238

5 SYNCHRONIZATION 241

5.1 CLOCK SYNCHRONIZATION 242

5.1.1 Physical Clocks 243

5.1.2 Clock Synchronization Algorithms 246

5.1.3 Use of Synchronized Clocks 251

5.2 LOGICAL CLOCKS 252

5.2.1 Lamport timestamps 252

5.2.2 Vector timestamps 256

5.3 GLOBAL STATE 258

5.4 ELECTION ALGORITHMS 262

5.4.1 The Bully Algorithm 262

5.4.2 A Ring Algorithm 263

5.5.1 A Centralized Algorithm 265

5.5 MUTUAL EXCLUSION 265

5.5.2 A Distributed Algorithm 266

5.5.3 A Token Ring Algorithm 269

5.5.4 A Comparison of the Three Algorithms 270

5.6 DISTRIBUTED TRANSACTIONS 271

5.6.1 The Transaction Model 272

5.6.2 Classification of Transactions 275

5.6.3 Implementation 278

5.6.4 Concurrency Control 280

5.7 SUMMARY 288

6 CONSISTENCY AND REPLICATION 291

6.1 INTRODUCTION 292

6.1.1 Reasons for Replication 292

6.1.2 Object Replication 293

6.1.3 Replication as Scaling Technique 296

6.2 DATA-CENTRIC CONSISTENCY MODELS 297

6.2.1 Strict Consistency 298

6.2.2 Linearizability and Sequential Consistency 300

6.2.3 Causal Consistency 305

6.2.4 FIFO Consistency 306

6.2.5 Weak Consistency 308

6.2.6 Release Consistency 310

6.2.7 Entry Consistency 313

6.2.8 Summary of Consistency Models 315

6.3 CLIENT-CENTRIC CONSISTENCY MODELS 316

6.3.1 Eventual Consistency 317

6.3.2 Monotonic Reads 319

6.3.3 Monotonic Writes 320

6.3.4 Read Your Writes 322

6.3.5 Writes Follow Reads 323

6.3.6 Implementation 324

6.4 DISTRIBUTION PROTOCOLS 326

6.4.1 Replica Placement 326

6.4.2 Update Propagation 330

6.4.3 Epidemic Protocols 334

6.5 CONSISTENCY PROTOCOLS 337

6.5.1 Primary-Based Protocols 337

6.5.2 Replicated-Write Protocols 341

6.5.3 Cache-Coherence Protocols 345

6.6 EXAMPLES 346

6.6.1 Orca 347

6.6.2 Causally-Consistent Lazy Replication 352

6.7 SUMMARY 357

7 FAULT TOLERANCE 361

7.1 INTRODUCTION TO FAULT TOLERANCE 362

7.1.1 Basic Concepts 362

7.1.2 Failure Models 364

7.1.3 Failure Masking by Redundancy 366

7.2 PROCESS RESILIENCE 368

7.2.1 Design Issues 368

7.2.2 Failure Masking and Replication 370

7.2.3 Agreement in Faulty Systems 371

7.3 RELIABLE CLIENT-SERVER COMMUNICATION 375

7.3.1 Point-to-Point Communication 375

7.3.2 RPC Semantics in the Presence of Failures 375

7.4 RELIABLE GROUP COMMUNICATION 381

7.4.1 Basic Reliable-Multicasting Schemes 381

7.4.2 Scalability in Reliable Multicasting 383

7.4.3 Atomic Multicast 386

7.5 DISTRIBUTED COMMIT 393

7.5.1 Two-Phase Commit 393

7.5.2 Three-Phase Commit 399

7.6 RECOVERY 401

7.6.1 Introduction 401

7.6.2 Checkpointing 404

7.6.3 Message Logging 407

7.7 SUMMARY 410

8 SECURITY 413

8.1 INTRODUCTION TO SECURITY 414

8.1.1 Security Threats,Policies,and Mechanisms 414

8.1.2 Design Issues 420

8.1.3 Cryptography 425

8.2 SECURE CHANNELS 432

8.2.1 Authentication 433

8.2.2 Message Integrity and Confidentiality 441

8.2.3 Secure Group Communication 444

8.3 ACCESS CONTROL 447

8.3.1 General Issues in Access Control 447

8.3.2 Firewalls 451

8.3.3 Secure Mobile Code 453

8.4 SECURITY MANAGEMENT 460

8.4.1 Key Management 461

8.4.2 Secure Group Management 465

8.4.3 Authorization Management 466

8.5 EXAMPLE:KERBEROS 472

8.6 EXAMPLE:SESAME 473

8.6.1 SESAME Components 474

8.6.2 Privilege Attribute Certificates(PACs) 477

8.7 EXAMPLE:ELECTRONIC PAYMENT SYSTEMS 478

8.7.1 Electronic Payment Systems 478

8.7.2 Security in Electronic Payment Systems 480

8.7.3 Example Protocols 484

8.8 SUMMARY 488

9 DISTRIBUTED OBJECT-BASED SYSTEMS 493

9.1 CORBA 494

9.1.1 Overview of CORBA 495

9.1.2 Communication 501

9.1.3 Processes 508

9.1.4 Naming 514

9.1.5 Synchronization 518

9.1.6 Caching and Replication 518

9.1.7 Fault Tolerance 520

9.1.8 Security 522

9.2 DISTRIBUTED COM 525

9.2.1 Overview of DCOM 526

9.2.2 Communication 531

9.2.3 Processes 534

9.2.4 Naming 537

9.2.5 Synchronization 541

9.2.6 Replication 541

9.2.7 Fault Tolerance 541

9.2.8 Security 542

9.3 GLOBE 545

9.3.1 Overview of Globe 545

9.3.2 Communication 553

9.3.3 Processes 554

9.3.4 Naming 557

9.3.5 Synchronization 559

9.3.6 Replication 560

9.3.7 Fault Tolerance 563

9.3.8 Security 563

9.4 COMPARISON OF CORBA,DCOM,AND GLOBE 565

9.4.1 Philosophy 566

9.4.2 Communication 567

9.4.3 Processes 567

9.4.4 Naming 568

9.4.5 Synchronization 569

9.4.6 Caching and Replication 569

9.4.7 Fault Tolerance 570

9.4.8 Security 570

9.5 SUMMARY 572

10 DISTRIBUTED FILE SYSTEMS 575

10.1 SUN NETWORK FILE SYSTEM 576

10.1.1 Overview of NFS 576

10.1.2 Communication 581

10.1.3 Processes 582

10.1.4 Naming 583

10.1.5 Synchronization 590

10.1.6 Caching and Replication 594

10.1.7 Fault Tolerance 597

10.1.8 Security 600

10.2.1 Overview of Coda 604

10.2 THE CODA FILE SYSTEM 604

10.2.2 Communication 606

10.2.3 Processes 608

10.2.4 Naming 609

10.2.5 Synchronization 610

10.2.6 Caching and Replication 615

10.2.7 Fault Tolerance 618

10.2.8 Security 620

10.3.1 Plan 9:Resources Unified to Files 623

10.3 OTHER DISTRIBUTED FILE SYSTEMS 623

10.3.2 XFS:Serverless File System 629

10.3.3 SFS:Scalable Security 635

10.4 COMPARISON OF DISTRIBUTED FILE SYSTEMS 638

10.4.1 Philosophy 638

10.4.2 Communication 639

10.4.3 Processes 639

10.4.4 Naming 640

10.4.5 Synchronization 641

10.4.6 Caching and Replication 641

10.4.8 Security 642

10.4.7 Fault Tolerance 642

10.5 SUMMARY 643

11 DISTRIBUTED DOCUMENT-BASED SYSTEMS 647

11.1 THE WORLD WIDE WEB 648

11.1.1 Overview of WWW 648

11.1.2 Communication 657

11.1.3 Processes 662

11.1.4 Naming 668

11.1.5 Synchronization 671

11.1.6 Caching and Replication 672

11.1.7 Fault Tolerance 676

11.1.8 Security 676

11.2 LOTUS NOTES 678

11.2.1 Overview of Lotus Notes 678

11.2.2 Communication 680

11.2.3 Processes 681

11.2.4 Naming 683

11.2.5 Synchronization 685

11.2.6 Replication 685

11.2.8 Security 688

11.2.7 Fault Tolerance 688

11.3 COMPARISON OF WWW AND LOTUS NOTES 691

11.4 SUMMARY 695

12 DISTRIBUTED COORDINATION-BASED SYSTEMS 699

12.1 INTRODUCTION TO COORDINATION MODELS 700

12.2 TIB/RENDEZVOUS 702

12.2.1 Overview of TIB/Rendezvous 702

12.2.2 Communication 704

12.2.3 Processes 708

12.2.4 Naming 709

12.2.5 Synchronization 710

12.2.6 Caching and Replication 712

12.2.7 Fault Tolerance 713

12.2.8 Security 715

12.3 JINI 716

12.3.1 Overview of Jini 717

12.3.2 Communication 719

12.3.3 Processes 721

12.3.4 Naming 724

12.3.5 Synchronization 727

12.3.7 Fault Tolerance 728

12.3.6 Caching and Replication 728

12.3.8 Security 729

12.4 COMPARISON OF TIB/RENDEZVOUS AND JINI 730

12.5 SUMMARY 733

13 READING LIST AND BIBLIOGRAPHY 737

13.1 SUGGESTIONS FOR FURTHER READING 737

13.1.1 Introduction and General Works 737

13.1.2 Communication 739

13.1.3 Processes 739

13.1.4 Naming 740

13.1.5 Synchronization 741

13.1.6 Consistency and Replication 742

13.1.7 Fault Tolerance 743

13.1.8 Security 744

13.1.9 Distributed Object-Based Systems 745

13.1.10 Distributed File Systems 746

13.1.11 Distributed Document-Based Systems 747

13.1.12 Distributed Coordination-Based Systems 748

13.2 ALPHABETICAL BIBLIOGRAPHY 749

INDEX 785