目　錄
前言
**部分　概念
第1章　為什么使用NoSQL3
1.1　關(guān)系型數(shù)據(jù)庫(kù)的價(jià)值3
1.1.1　獲取持久化數(shù)據(jù)3
1.1.2　并發(fā)4
1.1.3　集成4
1.1.4　近乎標(biāo)準(zhǔn)的模型4
1.2　阻抗失諧5
1.3　“應(yīng)用程序數(shù)據(jù)庫(kù)”與“集成數(shù)據(jù)庫(kù)”6
1.4　蜂擁而來的集群8
1.5　NoSQL登場(chǎng)9
1.6　要點(diǎn)12
第2章　聚合數(shù)據(jù)模型13
2.1　聚合14
2.1.1　關(guān)系模型與聚合模型示例14
2.1.2　面向聚合的影響19
2.2　鍵值數(shù)據(jù)模型與文檔數(shù)據(jù)模型20
2.3　列族存儲(chǔ)21
2.4　面向聚合數(shù)據(jù)庫(kù)總結(jié)23
2.5　延伸閱讀24
2.6　要點(diǎn)24
第3章　數(shù)據(jù)模型詳解25
3.1　關(guān)系25
3.2　圖數(shù)據(jù)庫(kù)26
3.3　無模式數(shù)據(jù)庫(kù)28
3.4　物化視圖30
3.5　構(gòu)建數(shù)據(jù)存取模型31
3.6　要點(diǎn)36
第4章　分布式模型37
4.1　單一服務(wù)器37
4.2　分片38
4.3　主從復(fù)制40
4.4　對(duì)等復(fù)制42
4.5　結(jié)合“分片”與“復(fù)制”技術(shù)43
4.6　要點(diǎn)44
第5章　一致性47
5.1　更新一致性47
5.2　讀取一致性49
5.3　放寬“一致性”約束52
5.4　放寬“持久性”約束56
5.5　仲裁57
5.6　延伸閱讀59
5.7　要點(diǎn)59
第6章　版本戳61
6.1　“商業(yè)事務(wù)”與“系統(tǒng)事務(wù)”61
6.2　在多節(jié)點(diǎn)環(huán)境中生成版本戳63
6.3　要點(diǎn)65
第7章　映射-化簡(jiǎn)67
7.1　基本“映射-化簡(jiǎn)”68
7.2　分區(qū)與歸并69
7.3　組合“映射-化簡(jiǎn)”計(jì)算72
7.3.1　舉例說明兩階段“映射-化簡(jiǎn)”73
7.3.2　增量式“映射-化簡(jiǎn)”76
7.4　延伸閱讀77
7.5　要點(diǎn)77
第二部分　實(shí)現(xiàn)
第8章　鍵值數(shù)據(jù)庫(kù)81
8.1　何謂“鍵值數(shù)據(jù)庫(kù)”81
8.2　鍵值數(shù)據(jù)庫(kù)特性83
8.2.1　一致性83
8.2.2　事務(wù)84
8.2.3　查詢功能84
8.2.4　數(shù)據(jù)結(jié)構(gòu)86
8.2.5　可擴(kuò)展性86
8.3　適用案例87
8.3.1　存放會(huì)話信息87
8.3.2　用戶配置信息87
8.3.3　購(gòu)物車數(shù)據(jù)87
8.4　不適用場(chǎng)合87
8.4.1　數(shù)據(jù)間關(guān)系87
8.4.2　含有多項(xiàng)操作的事務(wù)88
8.4.3　查詢數(shù)據(jù)88
8.4.4　操作關(guān)鍵字集合88
第9章　文檔數(shù)據(jù)庫(kù)89
9.1　何謂文檔數(shù)據(jù)庫(kù)90
9.2　特性91
9.2.1　一致性91
9.2.2　事務(wù)92
9.2.3　可用性93
9.2.4　查詢功能94
9.2.5　可擴(kuò)展性95
9.3　適用案例97
9.3.1　事件記錄97
9.3.2　內(nèi)容管理系統(tǒng)及博客平臺(tái)98
9.3.3　網(wǎng)站分析與實(shí)時(shí)分析98
9.3.4　電子商務(wù)應(yīng)用程序98
9.4　不適用場(chǎng)合98
9.4.1　包含多項(xiàng)操作的復(fù)雜事務(wù)98
9.4.2　查詢持續(xù)變化的聚合結(jié)構(gòu)98
第10章　列族數(shù)據(jù)庫(kù)99
10.1　何謂列族數(shù)據(jù)庫(kù)99
10.2　特性100
10.2.1　一致性103
10.2.2　事務(wù)104
10.2.3　可用性104
10.2.4　查詢功能105
10.2.5　可擴(kuò)展性107
10.3　適用案例107
10.3.1　事件記錄107
10.3.2　內(nèi)容管理系統(tǒng)與博客平臺(tái)108
10.3.3　計(jì)數(shù)器108
10.3.4　限期使用108
10.4　不適用場(chǎng)合109
第11章　圖數(shù)據(jù)庫(kù)111
11.1　何謂圖數(shù)據(jù)庫(kù)111
11.2　特性113
11.2.1　一致性114
11.2.2　事務(wù)114
11.2.3　可用性115
11.2.4　查詢功能115
11.2.5　可擴(kuò)展性119
11.3　適用案例120
11.3.1　互聯(lián)數(shù)據(jù)120
11.3.2　安排運(yùn)輸路線、分派貨物和基于位置的服務(wù)120
11.3.3　推薦引擎121
11.4　不適用場(chǎng)合121
第12章　模式遷移123
12.1　模式變更123
12.2　變更關(guān)系型數(shù)據(jù)庫(kù)的模式123
12.2.1　遷移全新項(xiàng)目124
12.2.2　遷移既有項(xiàng)目126
12.3　變更NoSQL數(shù)據(jù)庫(kù)的模式128
12.3.1　增量遷移130
12.3.2　遷移圖數(shù)據(jù)庫(kù)的模式131
12.3.3　改變聚合結(jié)構(gòu)132
12.4　延伸閱讀132
12.5　要點(diǎn)132
第13章　混合持久化133
13.1　各異的數(shù)據(jù)存儲(chǔ)需求133
13.2　混用各類數(shù)據(jù)庫(kù)134
13.3　將直接數(shù)據(jù)庫(kù)操作封裝為服務(wù)136
13.4　擴(kuò)展數(shù)據(jù)庫(kù)以增強(qiáng)其功能136
13.5　選用合適的數(shù)據(jù)庫(kù)技術(shù)138
13.6　企業(yè)使用混合持久化技術(shù)時(shí)的考量138
13.7　部署復(fù)雜度139
13.8　要點(diǎn)140
第14章　超越NoSQL141
14.1　文件系統(tǒng)141
14.2　事件溯源142
14.3　內(nèi)存映像144
14.4　版本控制145
14.5　XML數(shù)據(jù)庫(kù)145
14.6　對(duì)象數(shù)據(jù)庫(kù)146
14.7　要點(diǎn)146
第15章　選擇合適的數(shù)據(jù)庫(kù)147
15.1　程序員的工作效率149
15.2　數(shù)據(jù)訪問性能150
15.3　繼續(xù)沿用默認(rèn)的關(guān)系型數(shù)據(jù)庫(kù)150
15.4　抽離數(shù)據(jù)庫(kù)策略以降低風(fēng)險(xiǎn)151
15.5　要點(diǎn)152
15.6　結(jié)語153
參考資料157



Contents

Part I: Understand . 1
Chapter 1: Why NoSQL? 3
1.1 The Value of Relational Databases 3
1.1.1 Getting at Persistent Data 3
1.1.2 Concurrency 4
1.1.3 Integration . 4
1.1.4 A (Mostly) Standard Model 4
1.2 Impedance Mismatch 5
1.3 Application and Integration Databases 6
1.4 Attack of the Clusters 8
1.5 The Emergence of NoSQL . 9
1.6 Key Points 12
Chapter 2: Aggregate Data Models 13
2.1 Aggregates . 14
2.1.1 Example of Relations and Aggregates 14
2.1.2 Consequences of Aggregate Orientation 19
2.2 Key-Value and Document Data Models . 20
2.3 Column-Family Stores . 21
2.4 Summarizing Aggregate-Oriented Databases . 23
2.5 Further Reading . 24
2.6 Key Points 24
Chapter 3: More Details on Data Models . 25
3.1 Relationships . 25
3.2 Graph Databases . 26
3.3 Schemaless Databases 28
3.4 Materialized Views 30
3.5 Modeling for Data Access 31
3.6 Key Points 36
Chapter 4: Distribution Models . 37
4.1 Single Server 37
4.2 Sharding 38
4.3 Master-Slave Replication . 40
4.4 Peer-to-Peer Replication . 42
4.5 Combining Sharding and Replication 43
4.6 Key Points 44
Chapter 5: Consistency 47
5.1 Update Consistency . 47
5.2 Read Consistency . 49
5.3 Relaxing Consistency . 52
5.3.1 The CAP Theorem 53
5.4 Relaxing Durability . 56
5.5 Quorums . 57
5.6 Further Reading . 59
5.7 Key Points 59
Chapter 6: Version Stamps . 61
6.1 Business and System Transactions 61
6.2 Version Stamps on Multiple Nodes 63
6.3 Key Points 65
Chapter 7: Map-Reduce . 67
7.1 Basic Map-Reduce . 68
7.2 Partitioning and Combining . 69
7.3 Composing Map-Reduce Calculations . 72
7.3.1 A Two Stage Map-Reduce Example 73
7.3.2 Incremental Map-Reduce . 76
7.4 Further Reading . 77
7.5 Key Points 77
Part II: Implement 79
Chapter 8: Key-Value Databases 81
8.1 What Is a Key-Value Store . 81
8.2 Key-Value Store Features . 83
8.2.1 Consistency . 83
8.2.2 Transactions 84
8.2.3 Query Features . 84
8.2.4 Structure of Data 86
8.2.5 Scaling 86
8.3 Suitable Use Cases . 87
8.3.1 Storing Session Information . 87
8.3.2 User Pro.les, Preferences . 87
8.3.3 Shopping Cart Data 87
8.4 When Not to Use . 87
8.4.1 Relationships among Data . 87
8.4.2 Multioperation Transactions 88
8.4.3 Query by Data . 88
8.4.4 Operations by Sets 88
Chapter 9: Document Databases 89
9.1 What Is a Document Database? . 90
9.2 Features . 91
9.2.1 Consistency . 91
9.2.2 Transactions 92
9.2.3 Availability 93
9.2.4 Query Features . 94
9.2.5 Scaling 95
9.3 Suitable Use Cases . 97
9.3.1 Event Logging 97
9.3.2 Content Management Systems, Blogging Platforms . 98
9.3.3 Web Analytics or Real-Time Analytics . 98
9.3.4 E-Commerce Applications . 98
9.4 When Not to Use . 98
9.4.1 Complex Transactions Spanning Different Operations 98
9.4.2 Queries against Varying Aggregate Structure 98
Chapter 10: Column-Family Stores 99
10.1 What Is a Column-Family Data Store? . 99
10.2 Features . 100
10.2.1 Consistency 103
10.2.2 Transactions 104
10.2.3 Availability 104
10.2.4 Query Features . 105
10.2.5 Scaling 107
10.3 Suitable Use Cases 107
10.3.1 Event Logging 107
10.3.2 Content Management Systems, Blogging Platforms . 108
10.3.3 Counters 108
10.3.4 Expiring Usage . 108
10.4 When Not to Use . 109
Chapter 11: Graph Databases . 111
11.1 What Is a Graph Database? . 111
11.2 Features . 113
11.2.1 Consistency 114
11.2.2 Transactions 114
11.2.3 Availability 115
11.2.4 Query Features . 115
11.2.5 Scaling 119
11.3 Suitable Use Cases 120
11.3.1 Connected Data 120
11.3.2 Routing, Dispatch, and Location-Based Services . 120
11.3.3 Recommendation Engines 121
11.4 When Not to Use . 121
Chapter 12: Schema Migrations . 123
12.1 Schema Changes 123
12.2 Schema Changes in RDBMS 123
12.2.1 Migrations for Green Field Projects . 124
12.2.2 Migrations in Legacy Projects . 126
12.3 Schema Changes in a NoSQL Data Store . 128
12.3.1 Incremental Migration . 130
12.3.2 Migrations in Graph Databases 131
12.3.3 Changing Aggregate Structure 132
12.4 Further Reading . 132
12.5 Key Points 132
Chapter 13: Polyglot Persistence . 133
13.1 Disparate Data Storage Needs 133
13.2 Polyglot Data Store Usage 134
13.3 Service Usage over Direct Data Store Usage . 136
13.4 Expanding for Better Functionality 136
13.5 Choosing the Right Technology . 138
13.6 Enterprise Concerns with Polyglot Persistence . 138
13.7 Deployment Complexity 139
13.8 Key Points 140
Chapter 14: Beyond NoSQL 141
14.1 File Systems 141
14.2 Event Sourcing . 142
14.3 Memory Image . 144
14.4 Version Control . 145
14.5 XML Databases . 145
14.6 Object Databases . 146
14.7 Key Points 146
Chapter 15: Choosing Your Database 147
15.1 Programmer Productivity . 147
15.2 Data-Access Performance . 149
15.3 Sticking with the Default 150
15.4 Hedging Your Bets . 150
15.5 Key Points 151
15.6 Final Thoughts . 152
Bibliography . 153