本书是一本非常适合作研究生数据库教学用的教材。其内容包括了分布式数据库所用的基本原理、方法、重要的算法介绍及部分系统的实际实现技巧。第2版还增加了一些新的内容如:用整整两章来介绍并行数据库和分布式对象数据库管理系统。本书具有以下特点:1.内容较为全面,系统地覆盖了分布式数据库的主要方面;2.反映了目前的一些研究成果如:数据仓库,万维网和数据库,基于推理的技术以及移动DBMS;3.叙述清楚,容易阅读。 本书在清华大学计算机系研究生教学中使用了6年,反映较好。

PREFACE TO THE SECOND EDITION
PREFACE TO THE FIRST EDITION
1 1 NTRODUCTION
1.1 DISTRIBUTED DATA PROCESSING
l.2 WHAT IS A DISTRIBUTED DATABASE SYSTEM?
1.3 PROMISES OF DDBSs
1.3.1 Transparent Management of Distributed and Replicated Data
1.3.2 Reliability Through Distributed Transactions
1.3.3 Improved Performance
1.3.4 Easier System Expansion
1.4 COMPLICATING FACTORS
1.5 PROBLEM AREAS
l.5.1 Distributed Database Design
1.5.2 Distributed Query Processing
l.5.3 Distributed Directory Management
1.5.4 Distributed Concurrency Control
1.5.5 Distributed Deadlock Management
1.5.6 Reliability of Distributed DBMS
1.5.7 Operating System Support
1.5.8 Heterogeneous Databases
1.5.9 Relationship among Problems
1.6 BIBLIOGRAPHIC NOTES

2 OVERVIEW OF RELATIONAL DBMS
2.1 RELATIONAL DATABASE CONCEPTS
2.2 NORMALIZATION
2.2.1 Dependency Structures
2.2.2 Normal Forms
2.3 INTEGMTY RULES
2.4 RELATIONAL DATA LANGUAGES
2.4.1 Relational Algebra
2.4.2 Relational Calculus
2.4.3 Interface with Programming Languages
2.5 RELATIONAL DBMS
2.6 BIBLJOCRAPHIC NOTES

3 REVIEW OF COMPUTER NETWORKS
3.1 DATA COMMUNICATION CONCEPTS
3.2 TYPES OF NETWORKS
3.2.1 Topology
3.2.2 Communication Schemes
3.2.3 Scale
3.3 PROTOCOL STANDARDS
3.4 BROADBAND NETWORKS
3.5 WIRELESS NETWORKS
3.6 INTERNET
3.7 CONCLUDING REMARKS
3.8 BIBLIOGRAPHIC NOTES

4 DISTRIBUTED DBMS ARCHITECTURE
4.1 DBMS STANDARDIZATION
4.2 ARCHITECTURAL MODELS FOR DISTRIBUTED DBMSs
4.2.l Autonomy
4.2.2 Distribution
4.2.3 Heterogeneity
4.2.4 Architectural Alternatives
4.3 DISTRIBUTED DBMS ARCHITECTURE
4.3.1 Client/Server Systems
4.3.2 Peer-to-Peer Distributed Systems
4.3.3 MDBS Architecture
4.4 GLOBAL DIRECTORY ISSUES
4.5 CONCLUSION
4.6 BIBLIOGRAPHIC NOTES

5 DISTRIBUTED DATABASE DESIGN
5.l ALTERNATIVE DESIGN STRATEGIES
5.1.1 Top-Down Design Process
5.1.2 Bottom-Up Design Process
5.2 DISTRIBUTION DESIGN ISSUES
5.2.l Reasons for Fragmentation
5.2.2 Fragmentation Alternatives
5.2.3 Degree of Fragmentation
5.2.4 Correctness Rules of Fragmentation
5.2.5 Allocation Alternatives
5.2.6 Information Requirements
5.3 FRAGMENTATION
5.3.1 Horizontal Fragmentation
5.3.2 Vertical Fragmentation
5.3.3 Hybrid Fragmentation
5.4 ALLOCATION
5.4.1 Allocation Problem
5.4.2 Information Requirements
5.4.3 Allocation Model
5.4.4 Solution Methods
5.5 CONCLUSION
5.6 BIBLIOGRAPHIC NOTES
5.7 EXERCISES

6 SEMANTIC DATA CONTROL
6.1 VIEW MANAGEMENT
6.1.1 Views in Centralized DBMSs
6.1.2 Updates through Views
6.1.3 Views in Distributed DBMSs
6.2 DATA SECURITY
6.2.l Centralized Authorization Control
6.2.2 Distributed Authorization Control
6.3 SEMANTIC INTEGRITY CONTROL
6.3.l Celltra1ized Semantic Integrity Control
6.3.2 Distributed Semantic Integrity Control
6.4 CONCLUSION
6.5 BIBLIOGRAPHIC NOTES
6.6 EXERCISES

7 OVERVIEW OF QUERY PROCESSING
7.l QUERY PROCESSING PROBLEM
7.2 OBJECTIVES OF QUERY PROCESSING
7.3 COMPLEXITY OF RELATIONAL ALGEBRA OPERATJONS
7.4 CHARACTERIZATION OF QUERY PROCESSORS
7.4.l Languages
7.4.2 Types of Optimization
7.4.3 Optimization Timing
7.4.4 Statistics
7.4.5 Decision Sites
7.4.6 Exploitation of -the Network Topology
7.4.7 Exploitation of Replicated Fragments
7.4.8 Use of Semijoins
7.5 LAYERS OF QUERY PROCESSING
7.5.l Query Decomposition
7.5.2 Data Localization
7.5.3 Global Query Optimization
7.5.4 Local Query Opti1nization
7.6 CONCLUSION
7.7 BIBLIOGRAPHIC NOTES

8 QUERY DECOMPOSITION AND DATA LOCALIZATION
8.1. QUERY DECOMPOSITION
8.l.1 Normalization
8.1.2 Analysis
8.1.3 Elimination of Redundancy
8.1.4 Rewriting
8.2 LOCALIZATION OF DISTRIBUTED DATA
8.2.1 Reduction for Primary Horizolltal Fragmentation
8.2.2 Reduction for Vertical Fragmentation
8.2.3 Reduction for Derived Fragmentation
8.2.4 Reduction for Hybrid Fragmentation
8.3 CONCLUSION
8.4 BIBLIOGRAPHIC NOTES
8.5 EXERCISES

9 OPTIMIZATION OF DISTRIBUTED QUERIES
9.1 QUERY OPTIMIZATION
9.1.1 Search Space
9.l.2 Search Strategy
9.1.3 Distributed Cost Model
9.2 CENTRALIZED QUERY OPTIMIZATION
9.2.l INGRES Algorithm
9.2.2 System R A1gorithm
9.3 JOIN ORDERING IN FRAGMENT QUERIES
9.3.1 Join Ordering
9.3.2 Semijoin Based Algorithms
9.3.3 Join versus Semi join
9.4 DISTRJBUTED QUERY OPTIMIZATION ALGORITHMS
9.4.1 Distributed INpRES Algorithm
9.4.2 R* Algorithm
9.4.3 SDD-1 Algorithm
9.5 CONCLUSION
9.6 BIBLIOGRAPHIC NOTES
9.7 EXERCISES

10 INTRODUCTION TO TRANSACTION MANAGEMENT
10.1 DEFINITION OF A TRANSACTION
10.1.1 Termination Conditions of Thansactions
l0.1.2 Characterization of Transactions
10.1.3 Formalization of the Transaction Concept
10.2 PROPERTIES OF TRANSACTIONS
10.2.1 Atomicity
10.2.2 Consistency
l0.2.3 Isolation
l0.2.4 Durabi1ity
l0.3 TYPES OF TRANSACTIONS
l0.3.1 Flat Thansactions
l0.3.2 Nested Transactions
10.3.3 Workflows
l0.4 ARCHITECTURE REVISITED
l0.5 CONCLUSION
10.6 BIBLIOGRAPHIC NOTES

11 DISTRIBUTED CONCURRENCY CONTROL
11.1 SERIALIZABILITY THEORY
l1.2 TAXONOMY OF CONCURRENCY CONTROL MECHANISMS
11.3 LOCKING-BASED CONCURRENCY CONTROL ALGORITHMS
11.3.l Centralized 2PL
11.3.2 Primary Copy 2PL
11.3.3 Distributed 2PL
11.4 TIMESTAMP-BASED CONCURRENCY CONTROL ALGORITHMS
11.4.l Basic TO Algorithm
11.4.2 Conservative TO Algorithm
11.4.3 Multiversion TO Algorithm
1l.5 OPTIMISTIC CONCURRENCY CONTROL ALGORITHMS
11.6 DEADLOCK MANAGEMENT
11.6.1 Deadlock Prefelltion
l1.6.2 Deadlock Avoidance
11.6.3 Deadlock Detection and Resolution
l1.7 "RELAXED" CONCURRENCY CONTROL
11.7.1 Non-Serializable Schedules
11.7.2 Nested Distributed nansactions
11.8 CONCLUSION
11.9 BIBLIOGRAPHIC NOTES
11.10 'EXERCISES

12 DISTRIBUTED DBMS RELIABILITY
12.1 RELIABlLITY CONCEPTS AND MEASURES
12.l.1 System, State, and Failure
12.1.2 Reliability and Availability
12.1.3 Mean Time between Failures/Mean Time to Repair
12.2 FAILURES AND FAULT TOLERANCE IN DISTRIBUTED SYSTEMS
12.2.1 Reasons for Failures
l2.2.2 Basic Fault Tolerance Approaches and Techniques
12.3 FAILURES IN DISTRIBUTED DBMS
12.3.l Thansaction Failures
12.3.2 Site (System) Failures
12.3.3 Media Failures
12.3.4 Communication Failures
12.4 LOCAL RELIABILITY PROTOCOLS
l2.4.l Architectural Considerations
12.4.2 Recovery Information
12.4.3 Execution of LRM Commands
12.4.4 Checkpoillting
l2.4.5 Handling Media Failures
12.5 DISTRIBUTED RELIABILITY PROTOCOLS
12.5.1 Components of Distributed Reliability Protocols
12.5.2 Two-Phase Commit Protocol
12.5.3 Variations of 2PC
12.6 DEALING WITH SITE FAILURES
12.6.1 Termination and Recovery Protocols for 2PC
12.6.2 Three-Phase Commit Protocol
12.7 NETWORK PARTITIONING
12.7.1 Centralized Protocols
12.7.2 Voting-based Protocols
12.7.3 Replication and Replica Contro1 Protocols
l2.7.4 Strict Replica Control Protocols
l2.7.5 Lazy Replication Protocols
12.8 ARCHITECTURAL CONSIDERATIONS
12.9 CONCLUSION
12.10 BIBLIOGRAPHIC NOTES
12.11 EXERCISES

13 MRALLEL DATABASE SYSTEMS
13.1 DATABASE SERVERS
13.1.1 Database Server Approach
13.1.2 Database Servers and Distributed Databases
13.2 PARALLEL ARCHITECTURES
13.2.l Objectives
13.2.2 Functional Aspects
13.2.3 Parallel System Architectures
13.3 PARALLEL DBMS TECHNIQUES
13.3.l Data Placement
13.3.2 Query Parallelism
13.3.3 Para1lel Data Processing
13.3.4 Parallel Query Optimization
13.4 PARALLEL EXECUTION PROBLEMS
l3.4.1 Initialization
l3.4.2 Illterferences and Convoy Effect
13.4.3 Load Balancing
13.5 PARALLEL EXECUTION FOR HIERARCHlCAL ARCHITECTURE
13.5.1 Problem Formulation
13.5.2 Basic Concepts
13.5.3 Load Balancing Strategy
13.5.4 Performance Evaluation
l3.6 CONCLUSION
13.7 BIBLIOGRAPHIC NOTES
13.8 EXERCISES

14 DISTmBUTED OBJECT DATABASE MANAGEMENT SYSTEMS
14.1 FUNDAMENTAL OBJECT CONCEPTS AND MODELS
14.l.1 Object
14.1.2 Abstract Data Types
14.l.3 Composition (Aggregation)
14.1.4 Class
14.l.5 Collection
l4.1.6 Subtyping and Inheritance
l4.2 OBJECT DISTRIBUTION DESIGN
l4.2.1 Horizontal Class Partitioning
14.2.2 Vertical Class Partitioning
14.2.3 Path Partitioning
l4.2.4 Class Partitioning Algorithms
14.2.5 Allocation
14.2.6 Replication
l4.3 ARCHITECTURAL ISSUES
14.3.1 Alternative Client/Server Architectures
14.3.2 Cache Consistency
14.4 OBJECT MANAGEMENT
l4.4.1 Object Identifier Management
14.4.2 Pointer Swizzling
l4.4.3 Object Migration
14.5 DISTRIBUTED OBJECT STORAGE
l4.6 OBJECT QUERY PROCESSING
14.6.l Object Query Processor Architectures
14.6.2 Query Processing Issues
l4.6.3 Query Execution
l4.7 TRANSACTION MANAGEMENT
l4.7.1 Correctness Criteria
l4.7.2 Thansaction Models and Object Structures
14.7.3 Thansactions Managemeflt in Object DBMSs
l4.7.4 ThansaCtions as Objects
14.8 CONCLUSION
l4.9 BIBLIOGRAPHIC NOTES
14.l0 EXERCISES

15 DATABASE INTEROPERABILITY
15.1 DATABASE INTEGRATION
15.1.1 Schema Trans1ation
15.1.2 Schema Integration
15.2 QUERY PROCESSING
15.2.1 Query Processing Layers in Distributed Multi--DBMSs
15.2.2 Query Optimization Issues
15.3 TRANSACTION MANAGEMENT
15.3.1 Thansaction and Computation Model
15.3.2 Multidatabase Concurrency Control
15.3.3 Multidatabase Recovery
15.4 OBJECT ORIENTATION AND INTEROPERABILITY
15.4.1 Object Management Architecture
15.4.2 CORBA and Database Interoperability
15.4.3 Distributed Component Object Model
15.4.4 COM/OLE and Database Interoperability
15.5 CONCLUSION
l5.6 BIBLIOGRAPHIC NOTES
15.7 EXERCISES

16 CURRENT ISSUES
16.1 DATA DELIVERY ALTERNATIVES
16.2 DATA WAREHOUSING
16.2.l Architectures
16.2.2 OLAP Data Model
16.2.3 OLAP Servers
16.2.4 Research Issues
l6.3 WORLD WIDE WEB
16.3.1 Architecture and Protocols
16.3.2 Database Access
l6.3.3 Semistructured Data
16.3.4 Architectures for Information Integration
16.3.5 Research Projects and Open Issues
16.4 PUSH-BASED TECHNOLOGIES
16.4.1 Delivery Schedule Generation
16.4.2 Client Cache Management
16.4.3 Propagating Updates
16.5 MOBILE DATABASES
16.5.1 Directory Management
16.5.2 Caching
16.5.3 Broadcast Data
16.5.4 Query Processing and Optimization
16.5.5 Thansaction Management
16.6 BIBLIOGRAPHIC NOTES

BIBLIOGRAPHY
SUBJECT INDEX