Front cover

Draft Document for Review October 16, 2009 2:07 pm SG24-7093-01

MySQL to DB2
Conversion Guide
Complete guide to convert MySQL
database and application to DB2

Application enrichment through

advanced DB2 features

Application conversion
with detailed examples

Whei-Jen Chen
Angela Carlson

ibm.com/redbooks
Draft Document for Review October 16, 2009 10:22 am 7093edno.fm

International Technical Support Organization

MySQL to DB2 Conversion Guide

October 2009

SG24-7093-01
7093edno.fm Draft Document for Review October 16, 2009 10:22 am

Note: Before using this information and the product it supports, read the information in
“Notices” on page ix.

Second Edition (October 2009)

This edition applies to DB2 9.7 for Linux, UNIX, and Windows, MySQL 5.1.

This document created or updated on October 16, 2009.

© Copyright International Business Machines Corporation 2009. All rights reserved.

Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP
Schedule Contract with IBM Corp.
Draft Document for Review October 16, 2009 10:25 am 7093TOC.fm

Contents

Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
The team who wrote this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Become a published author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Summary of changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

October 2009, Second Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

Executive summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix

Chapter 1. DB2 for Linux, UNIX, and Windows. . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Product overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 DB2 Data Server Editions for the production environment . . . . . . . . . 3
1.2.2 Products for accessing System z and System i host data . . . . . . . . . 7
1.2.3 DB2 for pervasive platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.4 Additional DB2 data server features . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 DB2 for Linux, UNIX, and Windows architecture. . . . . . . . . . . . . . . . . . . . . 8
1.3.1 DB2 9.7 threaded architecture and process model. . . . . . . . . . . . . . . 9
1.3.2 DB2 database objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.3 DB2 catalog. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.4 DB2 utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.5 DB2 database access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.5.1 DB2 clients and drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.5.2 Application access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.5.3 DB2 application programming interfaces (APIs) . . . . . . . . . . . . . . . . 26

Chapter 2. MySQL database. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

2.1 MySQL licensing overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.2 MySQL architecture overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.2.1 Database client and non-client utilities . . . . . . . . . . . . . . . . . . . . . . . 37
2.2.2 Database server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.3 MySQL design and SQL compliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.3.1 MySQL directory structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.3.2 MySQL storage engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.3.3 MySQL standard SQL compliance . . . . . . . . . . . . . . . . . . . . . . . . . . 47

© Copyright IBM Corp. 2009. All rights reserved. iii

7093TOC.fm Draft Document for Review October 16, 2009 10:25 am

2.4 MySQL utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

2.4.1 Overview of the MySQL server-side programs and utilities . . . . . . . 49
2.4.2 Overview of the MySQL client-side programs and utilities . . . . . . . . 50
2.5 MySQL application programming interfaces (API) . . . . . . . . . . . . . . . . . . 51

Chapter 3. Planning the conversion from MySQL to DB2 . . . . . . . . . . . . . 55

3.1 Conversion project planning overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.1.1 Benefits of converting to DB2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.1.2 IBM conversion support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.1.3 Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.2 Application assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3 System planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3.1 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.2 Hardware. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.3 Conversion tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.4 The conversion process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.4.1 Porting preparation and installation. . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.4.2 Database structure porting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.4.3 Data porting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.4.4 Application porting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.4.5 Basic administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.4.6 Testing and tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

Chapter 4. Conversion scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

4.1 Application structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.1.1 Application Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.2 Database structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.3 System environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

Chapter 5. Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.1 DB2 Express-C 9.7 on Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.1.1 System requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.1.2 Installation procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.1.3 Instance creation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.1.4 Client setup on Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.2 Other software product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.2.1 Apache2 installation with DB2 support . . . . . . . . . . . . . . . . . . . . . . 105
5.2.2 PHP installation with DB2 support . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.3 IBM Data Movement Tool installation and usage . . . . . . . . . . . . . . . . . . 111
5.3.1 IBM Data Movement Tool prerequisites . . . . . . . . . . . . . . . . . . . . . 112
5.3.2 IBM Data Movement Tool installation . . . . . . . . . . . . . . . . . . . . . . . 113

Chapter 6. Database conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

6.1 Data type mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

iv MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:25 am 7093TOC.fm

6.2 Data definition language differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

6.2.1 Database manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
6.2.2 Table manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.2.3 Index manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.2.4 Triggers manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.2.5 Procedures and function manipulation . . . . . . . . . . . . . . . . . . . . . . 137
6.3 Other considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
6.4 Converting database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6.4.1 Automatic conversion using porting tools . . . . . . . . . . . . . . . . . . . . 142
6.4.2 Manual conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
6.4.3 Metadata transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
6.5 Sample database conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.5.1 Converting database objects using the IBM Data Movement Tool . 148
6.5.2 Manual database object conversion and enhancements . . . . . . . . 158

Chapter 7. Data conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

7.1 Data porting considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
7.1.1 Data porting commands and tools . . . . . . . . . . . . . . . . . . . . . . . . . 168
7.1.2 Differences in data formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
7.1.3 Differences in the user account management. . . . . . . . . . . . . . . . . 176
7.2 Sample project: data porting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
7.2.1 Export user data from MySQL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
7.2.2 Map MySQL user data to db2 user data . . . . . . . . . . . . . . . . . . . . . 193
7.2.3 Create DB2 user . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
7.2.4 Export MySQL application data. . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
7.2.5 Convert MySQL application data to DB2 format . . . . . . . . . . . . . . . 197
7.2.6 Import application data into DB2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
7.2.7 Basic data checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

Chapter 8. Application conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

8.1 Data Manipulation Language differences and similarities . . . . . . . . . . . . 208
8.1.1 SELECT syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
8.1.2 JOIN syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
8.1.3 UNION syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
8.1.4 Subquery syntax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
8.1.5 Grouping, having, and ordering. . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
8.1.6 Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
8.1.7 Implicit casting of data types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
8.1.8 String concatenation and NULL values . . . . . . . . . . . . . . . . . . . . . . 218
8.1.9 Record deletion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
8.1.10 Built-in functions and operators. . . . . . . . . . . . . . . . . . . . . . . . . . . 223
8.2 Application source conversion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
8.2.1 Converting MySQL Perl applications to DB2 . . . . . . . . . . . . . . . . . 224

Contents v
7093TOC.fm Draft Document for Review October 16, 2009 10:25 am

8.2.2 Converting MySQL PHP applications to DB2 . . . . . . . . . . . . . . . . . 227

8.2.3 Converting MySQL Ruby on Rails applications to DB2 . . . . . . . . . 238
8.2.4 Converting MySQL Java applications to DB2 . . . . . . . . . . . . . . . . . 241
8.2.5 Converting MySQL C/C++ applications to DB2 . . . . . . . . . . . . . . . 248
8.2.6 Converting Connector/ODBC applications to DB2 . . . . . . . . . . . . . 257
8.2.7 Condition handling in DB2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
8.2.8 Special conversions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.3 Additional application considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
8.3.1 What is the purpose of locking? . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
8.3.2 Concurrency control and transaction isolation . . . . . . . . . . . . . . . . 272
8.3.3 DB2 isolation levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
8.3.4 Locking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
8.3.5 Specifying the isolation level in DB2 . . . . . . . . . . . . . . . . . . . . . . . . 277

Chapter 9. Database administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

9.1 Database configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
9.1.1 DB2 configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
9.2 Database recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
9.2.1 DB2 database recovery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
9.3 Database replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
9.4 Data movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
9.4.1 DB2 data movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
9.5 High availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
9.6 Autonomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
9.7 Workload management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
9.8 Database management tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
9.8.1 DB2 Control Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
9.8.2 IBM Optim and Data Studio Suite Overview . . . . . . . . . . . . . . . . . . 314

Chapter 10. Testing and tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321

10.1 Test planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
10.1.1 Principles of software tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
10.1.2 Test documentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
10.1.3 Test phases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
10.1.4 1 Time planning and time exposure . . . . . . . . . . . . . . . . . . . . . . . 323
10.2 Data checking techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
10.2.1 IMPORT/LOAD messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
10.2.2 Data checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
10.3 Code and application testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
10.3.1 Application code check . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
10.3.2 Security testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
10.4 Troubleshooting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
10.4.1 Interpreting DB2 informational messages . . . . . . . . . . . . . . . . . . . 332

vi MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:25 am 7093TOC.fm

10.4.2 DB2 Tools for Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . 333

10.4.3 DB2 diagnostic logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
10.4.4 DB2 support information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
10.4.5 Monitoring tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
10.4.6 Visual Explain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
10.5 Initial tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
10.5.1 Table spaces design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
10.5.2 Physical placement of database objects . . . . . . . . . . . . . . . . . . . . 357
10.5.3 Buffer pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
10.5.4 Large transactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
10.5.5 SQL execution plan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
10.5.6 Configuration Advisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
10.5.7 Design Advisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374

Chapter 11. Advanced DB2 features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

11.1 pureXML . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
11.2 Data compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
11.3 Partitioning features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
11.3.1 Database partitioning feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
11.3.2 Table partitioning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
11.3.3 Multidimensional clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
11.4 Materialized query tables (MQT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
11.5 User-defined data types (UDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393

Appendix A. Mapping MySQL built-in functions and operators . . . . . . . 397

A.1 Grouping related functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
A.2 String functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
A.3 Numeric functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
A.4 Date and time functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
A.5 Comparing operators and other functions. . . . . . . . . . . . . . . . . . . . . . . . 408

Appendix B. Sample code for user defined functions . . . . . . . . . . . . . . . 411

B.1 Sample code for BIT_AND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
B.2 Sample code for FORMAT function . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
B.3 Sample code for RPAD and LPAD functions . . . . . . . . . . . . . . . . . . . . . 414
B.4 Sample code for GREATEST function . . . . . . . . . . . . . . . . . . . . . . . . . . 420
B.5 Sample code for LEAST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
B.6 Sample code for BIT_COUNT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
B.7 Sample code for SUBSTRING_INDEX. . . . . . . . . . . . . . . . . . . . . . . . . . 430

Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
Other publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
Online resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434

Contents vii
7093TOC.fm Draft Document for Review October 16, 2009 10:25 am

How to get Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

Help from IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

viii MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093spec.fm

Notices

This information was developed for products and services offered in the U.S.A.

IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local IBM representative for information on the products and services currently available in your area.
Any reference to an IBM product, program, or service is not intended to state or imply that only that IBM
product, program, or service may be used. Any functionally equivalent product, program, or service that
does not infringe any IBM intellectual property right may be used instead. However, it is the user's
responsibility to evaluate and verify the operation of any non-IBM product, program, or service.

IBM may have patents or pending patent applications covering subject matter described in this document.
The furnishing of this document does not give you any license to these patents. You can send license
inquiries, in writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, NY 10504-1785 U.S.A.

The following paragraph does not apply to the United Kingdom or any other country where such
provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION
PROVIDES THIS PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR
IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer
of express or implied warranties in certain transactions, therefore, this statement may not apply to you.

This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may
make improvements and/or changes in the product(s) and/or the program(s) described in this publication at
any time without notice.

Any references in this information to non-IBM Web sites are provided for convenience only and do not in any
manner serve as an endorsement of those Web sites. The materials at those Web sites are not part of the
materials for this IBM product and use of those Web sites is at your own risk.

IBM may use or distribute any of the information you supply in any way it believes appropriate without
incurring any obligation to you.

Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm
the accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on
the capabilities of non-IBM products should be addressed to the suppliers of those products.

This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.

COPYRIGHT LICENSE:

This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs in
any form without payment to IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating platform for which the
sample programs are written. These examples have not been thoroughly tested under all conditions. IBM,
therefore, cannot guarantee or imply reliability, serviceability, or function of these programs.

© Copyright IBM Corp. 2009. All rights reserved. ix

7093spec.fm Draft Document for Review October 16, 2009 10:21 am

Trademarks
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business
Machines Corporation in the United States, other countries, or both. These and other IBM trademarked
terms are marked on their first occurrence in this information with the appropriate symbol (® or ™),
indicating US registered or common law trademarks owned by IBM at the time this information was
published. Such trademarks may also be registered or common law trademarks in other countries. A current
list of IBM trademarks is available on the Web at http://www.ibm.com/legal/copytrade.shtml

The following terms are trademarks of the International Business Machines Corporation in the United States,
other countries, or both:

1-2-3® i5/OS® Rational®

AIX® IBM® Redbooks®
AS/400® IMS™ Redbooks (logo) ®
ClearCase® Informix® System i®
DB2 Connect™ InfoSphere™ System p®
DB2 Universal Database™ iSeries® System z9®
DB2® OpenPower® System z®
developerWorks® Optim™ Tivoli®
Distributed Relational Database OS/390® UniData®
Architecture™ PartnerWorld® WebSphere®
DRDA® POWER® z/OS®
eServer™ pSeries® z9®
Everyplace® pureXML® zSeries®
HACMP™ Rational Rose®

The following terms are trademarks of other companies:

AMD, the AMD Arrow logo, and combinations thereof, are trademarks of Advanced Micro Devices, Inc.

Snapshot, and the NetApp logo are trademarks or registered trademarks of NetApp, Inc. in the U.S. and
other countries.

SUSE, the Novell logo, and the N logo are registered trademarks of Novell, Inc. in the United States and
other countries.

Oracle, JD Edwards, PeopleSoft, Siebel, and TopLink are registered trademarks of Oracle Corporation
and/or its affiliates.

Red Hat, and the Shadowman logo are trademarks or registered trademarks of Red Hat, Inc. in the U.S. and
other countries.

VMware, the VMware "boxes" logo and design are registered trademarks or trademarks of VMware, Inc. in
the United States and/or other jurisdictions.

EJB, Enterprise JavaBeans, J2EE, Java, Java runtime environment, JavaBeans, JavaServer, JDBC, JSP,
MySQL, Solaris, Sun, and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the
United States, other countries, or both.

Access, ActiveX, Expression, Microsoft, MS, SQL Server, Visual Basic, Visual Studio, Windows Mobile,
Windows, and the Windows logo are trademarks of Microsoft Corporation in the United States, other
countries, or both.

x MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093spec.fm

Intel Pentium, Intel Xeon, Intel, Itanium-based, Itanium, Pentium, Intel logo, Intel Inside logo, and Intel
Centrino logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United
States, other countries, or both.

UNIX is a registered trademark of The Open Group in the United States and other countries.

Linux is a trademark of Linus Torvalds in the United States, other countries, or both.

Other company, product, or service names may be trademarks or service marks of others.

Notices xi
7093spec.fm Draft Document for Review October 16, 2009 10:21 am

xii MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093pref.fm

Preface

Switching database vendors is often considered an exhausting challenge for

database administrator and developer. Complexity, total cost, and the risk of
downtime are often the reasons that restrain IT decision makers to start the
conversion project. The primary goal of this book is to show that, with the proper
planning and guidance, converting from MySQL™ to IBM® DB2® for Linux,
UNIX, and Windows not only feasible but straight forward.

If you picked up this book, you are most likely considering converting to DB2 and
are probably aware of some of the advantages of to converting to DB2 data
server. In this IBM Redbooks® publication we discuss in detail how you can take
advantage of this industry leading database server.

This book is an informative guide that describes how to convert the database
system from MySQL 5.1 to DB2 9.7 on Linux®, and the steps involved in
enabling the applications to use DB2 instead of MySQL.

This guide also presents the best practices in conversion strategy and planning,
conversion tools, porting steps, and practical conversion examples. It is intended
for technical staff involved in a MySQL to DB2 conversion project.

The team who wrote this book

This book was produced by a team of specialists from around the world working
at the International Technical Support Organization, San Jose Center.

Whei-Jen Chen is a Project Leader at the International Technical Support

Organization, San Jose Center. She has extensive experience in application
development, database design and modeling, and DB2 system administration.
Whei-Jen is an IBM Certified Solutions Expert in Database Administration and
Application Development, as well as an IBM Certified IT Specialist.

Angela Carlson is a Software Engineer with IBM Canada. She

has four years technical experience working in the IT industry
focusing on relational database technology and application
development. She has experienced in developing relational
database applications with PHP, Perl, and Java™. In Angela’s
current position she works closely with IBM business partners
and their database needs, to enable their solutions with DB2. She also
researches and develops competitive material on DB2 and MySQL. Angela holds

© Copyright IBM Corp. 2009. All rights reserved. xiii

7093pref.fm Draft Document for Review October 16, 2009 10:21 am

a bachelor's degree of Software Engineering Science from the University of

Western Ontario.

Thanks to the following people for their contributions to this project:

Boris Bialek
Program Director, Information Management Partner Technologies, IBM Canada

Irina Delidjakova
Information Management Emerging Partnerships and Technologies, IBM Canada

Vlad Barshai
Information Management Emerging Partnerships and Technologies, IBM Canada

Martin Schlegel
Information Management Partner Technologies, IBM Canada

Emma Jacob
International Technical Support Organization, San Jose Center

Thanks to the authors of the previous editions of this book.

򐂰 Authors of the first edition, MySQL to DB2 Conversion Guide, published in
May 2004, were:
Whei-Jen Chen, Andreas Blank, Michael Hoeller, Rakesh Midha, Klaus Subtil

Become a published author

Join us for a two- to six-week residency program! Help write a book dealing with
specific products or solutions, while getting hands-on experience with
leading-edge technologies. You will have the opportunity to team with IBM
technical professionals, Business Partners, and Clients.

Your efforts will help increase product acceptance and customer satisfaction. As
a bonus, you will develop a network of contacts in IBM development labs, and
increase your productivity and marketability.

Find out more about the residency program, browse the residency index, and
apply online at:
ibm.com/redbooks/residencies.html

xiv MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093pref.fm

Comments welcome
Your comments are important to us!

We want our books to be as helpful as possible. Send us your comments about

this book or other IBM Redbooks publications in one of the following ways:
򐂰 Use the online Contact us review Redbooks form found at:
ibm.com/redbooks
򐂰 Send your comments in an e-mail to:
redbooks@us.ibm.com
򐂰 Mail your comments to:
IBM Corporation, International Technical Support Organization
Dept. HYTD Mail Station P099
2455 South Road
Poughkeepsie, NY 12601-5400

Preface xv
7093pref.fm Draft Document for Review October 16, 2009 10:21 am

xvi MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093chang.fm

Summary of changes

This section describes the technical changes made in this edition of the book and
in previous editions. This edition may also include minor corrections and editorial
changes that are not identified.

Summary of Changes
for SG24-7093-01
for MySQL to DB2 Conversion Guide
as created or updated on October 16, 2009.

October 2009, Second Edition

This revision reflects the addition, deletion, or modification of new and changed
information described below.

New information
򐂰 DB2 feature and functions of DB2 for Linux, UNIX®, and Windows® Version
9, 9.5, and 9.7
򐂰 IBM Data Movement Tool
򐂰 MySQL 5.1 features

Changed information
򐂰 DB2 and MySQL features and functions
򐂰 Conversion scenarios and examples

© Copyright IBM Corp. 2009. All rights reserved. xvii

7093chang.fm Draft Document for Review October 16, 2009 10:21 am

xviii MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093chang.fm

Executive summary

This IBM Redbook describes how to migrate MySQL 5.1 to DB2 Version 9.7 on Linux
and enable your applications on DB2. To further ease your migration, this informative
guide will cover best practices in migration strategy and planning, as well as the
step-by-step directions, tools, and practical conversion. After completing this book, it
will be clear to the technical reader that a MySQL to DB2 migration is easy and
straight forward.

Potential IBM customers seek migration information because DB2 offers

performance and functional capabilities that the competition can't compare with. DB2
Express C, our lightweight community edition, is free to develop, deploy and
redistribute, and is designed to give the IT community a powerful alternative to the
open source or free databases currently available.

DB2 Express C offers the same high quality, reliable, scalable features that you would
expect from an IBM enterprise database at no charge. Fixed Term License support is
available as well, at a lower price than the competition. The decision to migrate
becomes simple when you consider that DB2 can be easily deployed in the
development stack, while offering many additional features and ease of use.

Enterprise class features aimed to lower the total cost of ownership can be found in
every edition of DB2. DB2 has powerful autonomics which make installation,
configuration, maintenance and administration virtually hands free. DB2 9.7's
compression features help companies manage rising energy costs and reduce
datacenter sprawl by reducing storage requirements and improving I/O efficiency.

IBM is committed to providing products to our customers that are powerful and
affordable. DB2 provides industry leading features, such as pureXML, Workload
Management, and Granular Security. Using DB2 pureXML®, make XML data
processing even faster, more flexible, and more reliable. Manage workloads with new
threshold, priority and OS integration features in DB2 9.7. Keep data secure from
internal and external threats using the unparalleled security control in DB2 9.7.

Start taking advantage of these exciting new features and help your business
manage costs and simplify application development. Migrate your database systems
and applications today and discover why DB2 9.7 is a smarter product for a smarter
planet.

Arvind Krishna
General Manager
IBM Information Management

Executive summary xix

7093chang.fm Draft Document for Review October 16, 2009 10:21 am

xx MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Chapter 1. DB2 for Linux, UNIX, and

Windows
The goal of this chapter is to give an overview of the DB2 database server, its
architecture, the tools and utilities available with the server and application
programming interfaces.

In this chapter, we cover

򐂰 DB2 product overview
򐂰 DB2 for Linux, UNIX, and Windows architecture
򐂰 DB2 utilities
򐂰 DB2 database access

© Copyright IBM Corp. 2009. All rights reserved. 1

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

1.1 Introduction
IBM has an extremely strong history of database innovation and has developed a
number of highly advanced data servers. It started in the 60's when IBM
developed the Information Management System (IMS™), which is a hierarchical
database management system. IMS was used to maintained inventory for the
Saturn V moon rocket and the Apollo space vehicle. In the 70's IBM invented the
Relational Model and the Structured Query Language (SQL). In the 80's IBM
introduced DB2 for the mainframe (DB2 for z/OS®). This was the first database
which used relational modeling and SQL. DB2 for distributed platforms (DB2 for
Linux, UNIX, and Windows) was introduce in the 90's. Since then IBM continues
to development on DB2 for both mainframe and distributed platforms. Although
relational data model has become more prevalent in the industry; IBM still
realizes that the hierarchical data model is important. Therefore in July 2006 IBM
launched the first hybrid (also known as multi-structured) data server.

The release of DB2 for Linux, UNIX, and Windows Version 9 (DB2 9) data server
brought the most exciting and innovative database features to the market; these
features further enhanced with the release of DB2 9.5 and 9.7. DB2 9 introduced
many important features for both database administrators and application
developers. These features included pureXML®, autonomics, table partitioning,
data compression, and label based access control. DB2 9.5 enhanced the
manageability of the DB2 data server by introducing the threaded engine, easier
integration with HADR, workload management, enhancements to autonomics
and more. As DB2 9.7 was released the focus is to provide unparalleled reliability
and scalability for the changing needs of your business. Therefore, DB2 9.7
introduces enhancements to Version 9 and Version 9.5 features, such as
enhancements to data compression, performance, workload management,
security, and application development.

When this book was written DB2 9.7 had just been released on June 2009. DB2
9.7 is the database version we use through the book. DB2 9.7 is a highly
scalable, easy to install and manage hybrid data server. DB2 was developed to
meet the demands of even the most critical database applications. DB2 provides
a highly adaptable database environment while optimizing data storage through
backups and deep data row compression. This is managed through various
autonomics capabilities, such as self-tuning memory management and automatic
storage. DB2 deep embedded capabilities allow for ubiquitous deployment in
user directories and administrative installations for any size server. In a single
database, DB2 provides native storage and processing of both transactional XML
data in a pre-parsed tree format and relational data using pureXML technology!

2 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

1.2 Product overview

DB2 for Linux, UNIX, and Windows spans the spectrum from products on
handheld devices to large clusters and mainframes (see Figure 1-1). More
detailed information can be obtained from the DB2 Web site at:
http://www.ibm.com/db2/

Figure 1-1 DB2 product overview

1.2.1 DB2 Data Server Editions for the production environment

DB2 provides packages for users based on their business needs. In DB2 9.5,
IBM introduced the common client concept for all IBM data management
products that provides equal engine functionality across system boundaries. This
means that DB2 scales up and down according to your needs. Be it a desktop or
a heavy duty enterprise server, DB2 has consistent behavior across any platform
providing an easy move regardless of the size limit. DB2 widens the boundaries
even further since all IBM data servers including Informix® are now sharing a
common API for SQL. Therefore, whether using JDBC™ on DB2 for z/OS,
Informix 11 on an embedded system or DB2 for Linux, UNIX and Windows, the
driver code remains the same. In fact all database connections to any IBM data
server can be made with the same common client. In addition, IBM Optim™ Data
Studio, provides a new cross-family tool for both application development and
database management underlining commitment across and beyond DB2.

Various DB2 package formats include:

򐂰 DB2 Personal Edition

Chapter 1. DB2 for Linux, UNIX, and Windows 3

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

DB2 Personal Edition (PE) provides a single user database engine ideal for
deployment to PC based users. The PE includes the ability for remote
management, pureXML feature, and SQL replication feature, making it the
perfect choice for deployment in occasionally connected or remote office
implementations that don't require multi-user capability, that is, point of sales
systems.
PE does not accept remote database requests, however, it contains DB2
client components and serves as a remote client to a DB2 Server. The DB2
Personal edition can also be used for connecting and managing other DB2
data servers on the network.
The personal edition includes most of the features included in DB2 Express
Edition and runs in either 32-bit or 64-bit Intel® or AMD™ workstations for
either Windows or Linux operating systems.
򐂰 DB2 Express-C
DB2 Express-C is the no-charge community version of the DB2 data server. It
is targeted towards developers and Independent Software Vendors (ISVs) to
allow the development and deployment of applications including the free
distribution of DB2 Express-C itself. All applications developed with this
version of DB2 can be moved to a higher edition of DB2 for Linux, UNIX and
Windows and even DB2 for z/OS without any application changes if using
common SQL API set of the DB2 family.
This version of DB2 is free for download and is therefore perfectly suited for
DB2 educators and students. DB2 Express-C does not restrict the database
size and can be used in a 64-bit memory model. The code is optimized to use
up to a maximum of 2 CPU cores and 2 GB of memory. No fixpack updates
are available for this edition, however, new versions of Express-C are updated
and freely available for download at any time.
While this version does not include all the features of higher editions of DB2,
such as storage optimization, replication services, or high availability, it comes
with the award-winning pureXML technology to leverage both relational data
by being able to natively store XML data in a single database. Some of these
features can be activated by purchasing the DB2 Express Fixed Term License
(FTL). Obtaining the FTL enables a 1 year 24x7 support plus the ability to use
high availability and disaster recovery.
At any point users of DB2 Express-C can receive advice on the IBM DB2
Express Forum, monitored by IBM DB2 developers, by accessing the
following link:
http://www.ibm.com/developerworks/forums/forum.jspa?forumID=805
DB2 Express-C runs on Windows or Linux for both Intel and AMD on 32-bit or
64-bit architecture as well as Linux on Power (IBM System p® and System
i®).

4 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

You can download DB2 Express-C from this link:

hhttp://www-01.ibm.com/software/data/db2/express/
򐂰 DB2 Express + Fixed Term License
The plus fixed term license (FTL) is a one-year license subscription with 24x7
IBM support to DB2 Express, providing service-level assurances for problem
resolution and offers Fix Pack updates tied to the usual DB2 Fix Pack
schedule.
FTL also raises the DB2 Express-C limitations optimized to run for 4 CPU
cores and 4 GB of memory. Additional features with the DB2 Express FTL
include High Availability Disaster Recovery (HADR) feature and full SQL
replication capability.
Licensing is quite easy with DB2 Express-C + FTL, since it requires only
purchasing one FTL subscription per data server no matter how it is used or
as explained before, its size. This explicitly applies also to an HADR setup,
where no distinction is made between cold, warm or hot standby usage.
DB2 Express-C with FTL runs on Windows or Linux for both Intel or AMD on
32-bit or 64-bit architecture as well as Linux on Power (IBM System p and
System i).
More information on DB2 Express + Fixed Term License is available at:
http://www-01.ibm.com/software/data/db2/express/support.html?S_TACT=105AGX2
8&S_CMP=DLMAIN
򐂰 DB2 Express
DB2 Express edition is specifically tailored for small and medium businesses
(SMBs). It is designed for independent software vendors who need an
easy-to-install database integrated into their application software solution. It is
a multi-user version that supports local and remote applications in
stand-alone and local area network (LAN) environments.
DB2 Express utilizes up to 4 GB of memory and can be installed on server
with up to 200 processor value units. For more information on processor value
units, visit:
http://www.ibm.com/developerworks/db2/library/techarticle/dm-0611zikopoulos
2/
Some key features and strengths of the Express Edition are simplified
deployment, SQL replication, backup compression, autonomic management
features, such as the Self Tuning Memory Manager (STMM), adaptive utility
throttling, automatic storage management, configuration and design advisors,
health and fault monitors, automated backup, pureXML, high availability and
more.
DB2 Express runs on Linux, Solaris™ x86 or Windows platforms.

Chapter 1. DB2 for Linux, UNIX, and Windows 5

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

򐂰 DB2 Workgroup
DB2 Workgroup Server Edition is used primarily for small business to
medium environments. It includes support for a per authorized user based or
a per processor based licensing model designed to provide an attractive price
point for smaller installations while still providing a fully functional data server
on a wider range of platforms compared to lower editions of DB2.
Workgroup Server now includes the High Availability Disaster Recovery
(HADR) feature, TSA MP and Online Table Reorganization. The following
feature packs are available:
– Query Optimization including Materialized Query Tables (cached tables)
– Multidimensional Clustering and Query Parallelism
– pureXML
This DB2 edition can be deployed on systems with up to 400 processor value
units and 16 GB of memory.
While the DB2 Express edition only runs on Windows, Linux, or Solaris the
Workgroup Server edition adds support for AIX®, HP-UX on Itanium64 and
Solaris on x86 and Sparc.
򐂰 DB2 Enterprise
DB2 Enterprise Server Edition meets the database server needs for any size
business. This product is the ideal foundation for building data warehouses,
transaction processing, or Web-based solutions, as well as a back-end for
packaged solutions like ERP, CRM, and SCM. In addition, the DB2 Enterprise
Server Edition offers connectivity and integration for other enterprise DB2 and
Informix data sources.
The Enterprise Server Edition does not pose limits to the maximum memory
or number of CPU cores. It can be licensed with either authorized user
licenses or processor value unit licenses.
In addition to features offered in the Workgroup Server Edition, the following
features are also available: Query Parallelism, Multidimensional Clustering,
Materialized Query Tables, Table Partitioning, and Connection Concentration.
With the DB2 feature pack the following can also be added: Performance
Optimization Feature, Advanced Access® Control Feature, Storage
Optimization Feature and Geodetic Data Management Feature.
DB2 Enterprise Server Edition runs on Windows (32 and 64-bit), Linux (Intel/
AMD 64-bit, System i, System p, System z®), AIX, Solaris (Sparc and x64)
and HP-UX (ia64).
򐂰 InfoSphere™ Warehouse
InfoSphere Warehouse (formerly known as DB2 Warehouse) is a powerful
platform for building business intelligence (BI) solutions. InfoSphere
Warehouse comes as single integrated software package, using DB2

6 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Enterprise Server Edition as its base. It provides an ideal solution for

companies that need to consolidate data marts, information silos and
business analytics.
InfoSphere Warehouse contains database management tools and embedded
data movement and transformation tools. In addition to the base feature set of
DB2 Enterprise Server Edition, the DB2 feature packs containing Data
Partitioning Feature (DPF) are standard in all editions of InfoSphere
Warehouse. Workload Management and Deep Data Row Compression are
optional features for InfoSphere Warehouse Enterprise Base Edition.
The overall integrated software package for InfoSphere Warehouse has
support for slice and dice analytics and provides direct support for optimized
OLAP analytics against the dynamic warehouse. Embedded analytical
features include capabilities for Data Mining, In-line Analytics, Cubing
Services and Unstructured Analytics.
򐂰 Other DB2 editions
There are also DB2 versions for System i and System z available, see the
following Web site for details:
http://www.ibm.com/software/data/db2/

1.2.2 Products for accessing System z and System i host data

With the following DB2 products, enterprise systems can be extended to access
the host data in a System z or System i system:
򐂰 DB2 Connect™ Personal Edition
The DB2 Connect Personal Edition provides the application programming
interface (API) drivers and connectivity infrastructure to enable direct
connectivity from desktop applications to System z and System i data servers.
This product is specifically designed and is licensed for enabling two-tier,
client-server applications running on individual workstations, and as such, is
not appropriate for use on servers.
򐂰 DB2 Connect Enterprise Edition
The DB2 Connect Enterprise Edition addresses the needs of organizations
that require robust connectivity from a variety of desktop systems to System i
and System z database servers. DB2 client software is deployed on desktop
systems, and provides drivers that connect client-server applications running
on these desktop systems to a DB2 Connect server (gateway) that accesses
host data. The licensing model for this product is user based.
򐂰 DB2 Connect Application Server Edition
Technically the DB2 Connect Application Server Edition product is identical
to the DB2 Connect Enterprise Server, however, its licensing terms and

Chapter 1. DB2 for Linux, UNIX, and Windows 7

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

conditions are meant to address specific needs of multi-tier, client-server

applications, as well as applications that utilize Web technologies. DB2
Connect Application Server Edition license charges are based on the size and
number of processors available to the application servers where the
application is running.
򐂰 DB2 Connect Unlimited Edition
The DB2 Connect Unlimited Edition product is ideal for organizations with
extensive usage of DB2 Connect, especially where multiple applications are
involved. This product provides program code of the DB2 Connect Personal
Edition as well as program code identical to the DB2 Connect Application
Server Edition for unlimited deployment throughout an organization.

1.2.3 DB2 for pervasive platforms

DB2 Everyplace® Edition supports a wide variety of handheld devices such as
Palm OS, Windows Mobile® for Pocket PC, Windows desktops, Symbian, QNX
Neutrino and various Linux distributions. Its strength is to synchronize with data
from other systems, leveraging security features including table-level encryption
and communication encryption such as SSL.

DB2 Everyplace is also extremely flexible by being completely self-managed and

by supporting many programmable API interfaces such as ODBC, JDBC, .NET,
DB2 CLI, and so on. Its engine is highly flexible, which makes it easy to move
DB2 Everyplace databases from one supported mobile device to another.

1.2.4 Additional DB2 data server features

In addition to the product offerings for Linux, UNIX, Windows, and accessing
System z and System i host data, DB2 also offers a great variety of features,
which are included in various DB2 packages. Information about additional DB2
features and products can be found at:
http://www.ibm.com/software/data/db2/9/features.html?S_CMP=wspace

1.3 DB2 for Linux, UNIX, and Windows architecture

Data servers provide software services for the secure and efficient management
of structured information. DB2 is a hybrid relational and XML data server. A data
server refers to a computer where the DB2 database engine is installed. The
DB2 engine is a full-function, robust database management system that includes
optimized SQL support based on actual database usage and tools to help

8 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

manage the data. In this section we will discuss the DB2 engine architecture and
the database objects used to maintain your database.

1.3.1 DB2 9.7 threaded architecture and process model

Figure 1-2 on page 9 shows the DB2 9.7 architecture overview. As of DB2 9.5 a
threaded model architecture is used. There are many advantages to using a
threaded model architecture. Firstly, a new thread requires less memory and
fewer operating system resources than a process, because some operating
system resources can be shared among all threads within the same process.
Moreover, on some platforms, the context switch time for threads is less than
that for processes, which can improve performance.

Figure 1-2 DB2 Threaded Architecture

From a client-server perspective, the client code and the server code are
separated into different address spaces. The application code runs in the client
process, while the server code runs in a separate process. The client process
can run on either the same machine as the data server or a different one,
accessing the data server through a programming interface. The memory units
are allocated for database managers, databases, and applications.

Since DB2 is running with a threaded architecture all threads within the engine
process share the same address space, meaning all threads can immediately

Chapter 1. DB2 for Linux, UNIX, and Windows 9

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

see new memory allocations. This creates a simplified memory model by

allowing memory growth and shrinkage though control of a single memory
parameter for an entire instance. This is automatically performed by the Self
Tuning Memory Manager, which can also tune other memory parameters for
best performance without DBA intervention. Its adaptive algorithm is able to react
to unforeseen memory requirements in DB2 caused by workloads running
against it.

To enable access to a specific database, the DB2 instance process responsible

for the database must be running on the DB2 server. When an instance process
is started, several processes are created which interact with one another to
maintain connected applications and the database. There are several
background processes in DB2 that are pre-started, others start on a need-only
basis. Several important background processes are:
򐂰 The main process is the DB2 System Controller (db2sysc), which runs the
entire DB2 engine infrastructure. The DB2 data server activities are
performed by Engine Dispatchable Units (EDU), which are defined as
threads running within one single operating system process.
򐂰 A second DB2 background process is started together with the system
controller and is called the DB2 Watch Dog (db2wdog). Its responsibility is to
watch and monitor the system controller and to react to error conditions.
򐂰 For autonomic tasks, another process is initialized when the database is
activated. That can happen either manually or triggered by a client
connection. We call this process the Autonomic Computing Daemon (db2acd)
and runs autonomic tasks such as health-monitoring, auto-runstats, and the
administration scheduler on the client side.
򐂰 There is also the db2fmp process, which de-coupled from the DB2 system
controller process serves thread-safe stored procedures and user defined
functions (UDFs).

1.3.2 DB2 database objects

In this section, we introduce the DB2 objects and their relationship to each other.
Figure 1-3 on page 11 shows the basic DB2 database objects.

10 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Figure 1-3 DB2 objects relationship

򐂰 Instances
A DB2 instance represents the database management system. It controls
how data is manipulated and manages system resources assigned to it. Each
instance is a complete, fairly independent environment containing all the
database partitions defined for a given parallel database system. An instance
can have its own set of databases (which other instances cannot access
directly), and all database partitions share the same system directories. Each
instance has separate security from other instances on the same machine
(system), allowing for situations where both production and development
environments are run on the same machine without interference. In order to
connect to a database, any database client must first establish a network
connection to the instance.
򐂰 Databases
A database is a structured collection of data, which is stored within tables.
Since DB2 9, data within tables, can be stored as both relational data and
XML documents natively in a pre-parsed tree format within a table column.
Each database includes a set of system catalog tables that describe the
logical and physical structure of the object in the database, a configuration file

Chapter 1. DB2 for Linux, UNIX, and Windows 11

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

containing the parameter values configured for the database, and a recovery
log. Figure 1-4 on page 12 shows the relationship between instances,
databases, and tables.

Figure 1-4 Relationship between instances, databases, and tables

򐂰 Database partition groups

A database partition group is a set of one or more database partitions
(Figure 1-5 on page 13). A database partition group must be created prior to
the creation of the tables in a database. This is where the table spaces will be
stored. Once the database partition group is there, a table space can be
created where tables will be stored. If a partition group is not specified, there
is a default group where table spaces are allocated. In a non-partitioned
environment, all the data resides in a single partition, therefore, it is
unnecessary to worry about partition groups in simple setups.

12 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Figure 1-5 Database partition groups in a database

򐂰 System catalog tables

Each database includes a set of system catalog tables, which describe the
logical and physical structure of the data. DB2 creates and maintains an
extensive set of system catalog tables for each database. These tables
contain information regarding definitions of database objects including user
tables, views, indexes, and security information about the privileges that users
have on these objects. Catalog tables are created when the database is
created, and are updated during normal operations. They cannot be explicitly
created or dropped; however, they can be queried for their contents using the
catalog views.
򐂰 Table spaces
A database is organized into subdivided table spaces which store data. When
creating a table, it can be decided to have certain objects such as indexes and
large object (LOB) data kept separately from the rest of the table data. A table
space is equally spread over one or more physical storage devices, this is
called striping. When using a database, if no additional table spaces are
created by the user, the default user table space is used. This means DB2
allows you as much control as needed.
When table spaces are created, they reside in database partition groups. The
table space definitions and attributes are maintained in the database system
catalog. Each table space has at least one container assigned to it. A
container is an allocation of physical storage such as a file or a device.
Table spaces come in two flavors, system managed space (SMS) or database
managed space (DMS), as shown in Figure 1-6 on page 14. In an SMS table

Chapter 1. DB2 for Linux, UNIX, and Windows 13

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

space, each container is a directory in the file system of the operating system.
This type of table spaces allows the operating system's file manager to
control the storage space. In a DMS table space, each container is either a
resizable file, or a pre-allocated physical device such as a disk which the
database manager must control.

Figure 1-6 DMS / SMS Table spaces and containers

When using SMS or DMS in combination with container files you can choose
how DB2 is handling these files. For example, choosing to enable various
optimization features if supported by the operating systems, that is, Direct I/O
(to bypass file system caching; always enabled with raw and block devices),
Vector I/O (reading contiguous data pages from disk into contiguous portions
of memory), and Async I/O (non-sequential processing of read and write
requests across multiple disks to avoid delays from synchronous events).
When using the Automatic Storage feature in DB2, one can simply specify
folders where the database can automatically create and manage DMS table
spaces. When more space is required the database manager automatically
allocates more space. Table spaces can be automatically resized using this
feature. This feature provides a convenient and worry free operation scenario.
Manual operation can be done without having to specifying container files.
򐂰 Containers
A container is a physical storage device. It can be identified by a directory
name, a device name, or a file name. A container is assigned to a table
space. A single table space can span many containers, but each container
can belong to only one table space.
򐂰 Buffer pools
A buffer pool is the amount of memory allocated to cache table and index
data pages. The purpose of the buffer pool is to improve system performance.

14 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Think of it as a database controlled file system cache. Data can be accessed

much faster from memory than from disk. Therefore, the fewer times the
database manager needs to read from or write to a disk (I/O) synchronously,
the better the performance of the application. The size of the buffer pool is the
single most important performance tuning area, to help reduce the delay
caused by synchronous I/O.
Buffer pool memory can be automatically tuned online in the same way most
other memory related parameters. The feature responsible for automatic
tuning is called the Self Tuning Memory Manager (STMM), which allocates
and releases new memory from the OS by shifting unused memory within
DB2 to components.
򐂰 Schemas
A schema is an identifier, by default the user ID, which qualifies tables and
other database objects. A schema can be owned by an individual, and the
owner can control access to the data and the objects within it. A schema
name is used as the first part of a two-part object name. For example, a
schema named Smith might qualify a table named SMITH.PAYROLL.
򐂰 Tables
A database presents data as a collection of tables. Data within a table is
arranged in columns and rows. A table can contain XML documents that are
natively stored as a parsed hierarchical format, as shown in Figure 1-7 on
page 16. The data in the table is logically related, and relationships can be
defined between tables. Table data is accessed by using Structured Query
Language (SQL) or XQuery with XPath expressions. Both are standardized
query languages for defining and manipulating both relational and XML data
in a database. A query is used in applications or by users to retrieve data from
a database. A typical query for relational data uses SQL to create a statement
in the form of:
SELECT <column_name> FROM <table_name>
A typical XQuery for table shown in figure 1-7 would look like this:
xquery
db2-fn:sqlquery("SELECT INFO FROM XMLEmployeeInfo
WHERE EmpID=1001")/customerinfo/name
This allows us to iterate through all our table rows and return customer
information from each XML document stored in each row for the example
table outlined in Figure 1-7 on page 16. DB2 and the XQuery language also
allows us to modify and update a sub tree of an XML document in place
without having to re-write the whole document. This is all possible, since we
already store XML documents in a pre-parsed and hierarchical format.
db2 create table dept (EmpID int,…, XMLEmployeeInfo xml)

Chapter 1. DB2 for Linux, UNIX, and Windows 15

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

Figure 1-7 Relational and XML data in a single table

When using the Deep Data Row Compression feature DB2 is able to
transparently compress and decompress table rows (for each table with
compression turned on). This feature can effectively save 45-80% space on
disk. Compressed rows in a table are compressed when pre-fetched to buffer
pool memory and left in a compressed state until they are actually being
used. Although decompression of the data when its fetched adds a slight
overhead, I/O bound workloads will have a performance gain. This is due to
the reduced amount of data we actually need to read and write from or to disk
as well as saved memory.
򐂰 Views
A view provides a different way of looking at data from one or more tables; it is
a named specification of a result table. The specification is a SELECT
statement that runs whenever the view is referenced in a SQL statement. A
view has columns and rows just like a base table. All views can be used just
like base tables for data retrieval. Figure 1-8 on page 17 shows the
relationship between tables and views.

16 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Figure 1-8 Relationship between tables and views

򐂰 Indexes
An index is a set of keys, each pointing to rows in a table. For example, table
A has an index based on the first column in the table (Figure 1-9 on page 18).
This key value provides a pointer to the rows in the table: value 19 points to
record KMP. If searching for this particular record, a full table scan can be
avoided since we have an index defined. Except for changes in performance,
users of this table are unaware that an index is being used. DB2 decides
whether to use the index or not. DB2 also provides tools, such as the Design
Advisor, that can help decide what indexes would be beneficial.
An index allows efficient access when selecting a subset of rows in a table by
creating a direct path to the data through pointers. The DB2 SQL Optimizer
chooses the most efficient way to access data in tables. The optimizer takes
indexes into consideration when determining the fastest access path.
Indexes have both benefits and disadvantages. One should be careful when
defining indexes and take into consideration costs associated with update,
delete, and insert operations and maintenance such as reorganization and
recovery.

Chapter 1. DB2 for Linux, UNIX, and Windows 17

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

Figure 1-9 Relationship between indexes and tables

1.3.3 DB2 catalog

In DB2, the metadata is stored in a set of base tables and views called the
catalog. The catalog contains information about the logical and physical structure
of the database objects, object privileges, integrity information, and more.

The catalog is automatically created with the database. The base tables are
owned by the SYSIBM schema and stored in the SYSCATSPACE table space. On
top of the base tables, the SYSCAT and SYSSTAT views are created. SYSCAT
views are the read-only views that contain the object information and are found in
the SYSCAT schema. SYSSTAT views are updateable views containing statistical
information that are found in the SYSTAT schema. The complete DB2 catalog
views can be found in DB2 SQL Reference Volume 1 and 2 available for
download under the following link:
http://www.ibm.com/support/docview.wss?rs=71&uid=swg27015148

1.4 DB2 utilities

All DB2 system commands are installed in the sqllib/bin directory during
installation. Some of the most important commands in DB2 are listed in
Table 1-1, Table 1-2, Table 1-3 on page 19, Table 1-4 on page 20 and Table 1-5
on page 20. You can find more information in the DB2 manuals, especially the
Command Reference available at:
http://www.ibm.com/support/docview.wss?rs=71&uid=swg27015148

18 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Table 1-1 DB2 instance commands

Command Description Example

db2start Start the default instance db2start

db2stop Stop the current instance db2stop -f

db2icrt Create an instance db2icrt -u db2fenc1 db2inst1

db2idrop Drop and instance db2idrop -f db2inst1

db2ilist List all instance db2ilist

db2imigr Convert an instance after upgrading db2imigr -u db2fenc1 db2inst1

DB2

db2iupdt Update an instance after installation of db2iupdt -u db2fenc1 db2inst1

a fix pack

Table 1-2 Handling general database tasks

Description Example

Deactivate a database db2 deactivate db mydb

View Database Manager Settings db2 get dbm cfg show detail

Change a Database Manager Setting db2 update dbm cfg using health_mon off

View Database Settings db2 get db cfg show detail

Change a Database Setting db2 update db cfg using
SELF_TUNING_MEM on

Viewing registry values db2set

Changing registry parameters db2set DB2AUTOSTART=yes

Viewing catalogued databases db2 list db directory

Viewing catalogued nodes db2 list node directory

List all connected applications db2 list applications all

Force applications off db2 force application (41408, 55458)

List utilities db2 list utilities

Get a database snapshot db2 get snapshot for database on mydb

Table 1-3 DB2 DAS instance commands

Command Description Example

Chapter 1. DB2 for Linux, UNIX, and Windows 19

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

db2admin Starting and stopping the DB2 Administration db2admin start

Server

dasauto Autostart DB2 Administration Server dasauto -on

dascrt Create a DB2 Administration Server dascrt -u dasusr1

dasdrop Remove a DB2 Administration Server dasdrop

dasmigr Convert a DB2 Administration Server dasmigr

Table 1-4 Informational commands

Command Description Example

db2level Shows the current Version and Service db2level

Level

db2look Extracts DDL statements db2look -d dep -a -e -o

db2look.sql

db2dart Database analysis and reporting tool db2dart dbaddr

db2pd Troubleshooting Tool db2pd -db sample -locks

Table 1-5 Graphical tools

Tool Command Purpose

IBM Data Studio Integrated development environment

package which includes all administrative
capabilities as well as an integrated Eclipse
development environment for Java, XML,
and Web services.

Optim Development A purchasable integrated development

Studio environment for advanced development of
DB2 databases

Optim Database A purchasable integrated development

Administrator environment for advanced administration of
DB2 databases

DB2 installer db2setup Installs DB2 and creates instances

DB2 instance installer db2isetup Creates instances

Control Center db2cc Administer instances, databases

Replication Center db2rc Administer replication between servers

Satellite Admin. Center db2cc Administer collections of DB2 servers

20 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Command Line db2 or Execute DB2 commands at the command

Processor (CLP) db2cmd line

Health Center db2hc View/resolve health monitor alerts

Task Center db2cc –tc Schedule, run tasks; notify contacts

Journal db2cc -j Monitor jobs, recovery history, and so on

Configuration Assistant db2ca Configure instances, databases

Note: The Control Center and its associated components have been
deprecated in Version 9.7 and might be removed in a future release. It is
recommended to use the new suite of GUI tools for managing DB2 data and
data-centric applications. These new tools include the IBM Data Studio, the
Optim Development Studio and the Optim Database Administrator

1.5 DB2 database access

IBM introduced common application development and tooling support for DB2 for
Linux, UNIX and Windows (as of 9.5), DB2 for z/OS and Informix Dynamic
Server. This generally lowers development costs through reuse and support of
common components across IBM Data Servers.

More information to application development, tools, clients and drivers can be

found here:
http://www.ibm.com/software/data/db2/ad/

and
http://www.ibm.com/developerworks/db2/library/techarticle/dm-0804zikopoulos

1.5.1 DB2 clients and drivers

DB2 clients and drivers are used to access databases that reside on DB2
servers. A database cannot be created on a DB2 client.

IBM Data Server clients

DB2 data server 9.7 comes with the following clients:
򐂰 IBM Data Server Runtime Client
򐂰 IBM Data Server Client

Chapter 1. DB2 for Linux, UNIX, and Windows 21

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

The IBM Data Server Runtime client offers the basic client functionality and
includes drivers for ODBC, CLI, ADO.NET, OLE DB, PHP, Ruby, Perl-DB2, JDBC
and SQLJ. This client already includes the drivers and the capabilities to define
data sources. Furthermore the Lightweight Directory Access Protocol (LDAP) is
available as well.

Additionally the IBM Data Server client provides vast amounts of sample code in
various languages, header files for application development and graphical
administration and development tools, such as the DB2 Control Center, the IBM
Data Studio, the MS® Visual Studio® Tools and more.

Figure 1-10 on page 22 illustrates how to connect to a DB2 data server using the
IBM Data Server clients.

Figure 1-10 DB2 9.7 client and drivers

IBM Optim Data Studio is a comprehensive suite of integrated Eclipse tools

geared towards both database developers and database administrators. It
reduces the time to perform day-to-day administration tasks, create, deploy, and
debug SQL and Java stored procedures, deploy data-centric Web services, and
create queries for relational and XML data using SQL and XQuery. It supports
multiple IBM data servers including DB2 for Linux, UNIX, and Windows, DB2 for
i5/OS® and z/OS, Apache Derby and the Informix Dynamic Server. Since it is

22 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

built on the extensible Eclipse framework this IDE includes a number of plug-ins
to support programming languages, such as Java, C/C++, PHP, Ruby, Perl, and
so on. Other plug-ins are available to maintain the written application sources in
various source code repositories for example the Concurrent Versions System
(CVS) or IBM Rational® ClearCase® from within IBM Optim Data Studio.

If running mixed versions of DB2 servers and clients it is good to know, that DB2
Clients from DB2 UDB Version 8, DB2 9.1 or 9.5 for Linux, UNIX, and Windows
are still supported and able to connect to a DB2 9.7 data server. In the reversed
direction the newer IBM Data Server clients from version 9.7 can also connect to
the earlier DB2 9.1 and DB2 UDB version 8 servers using the IBM Data Server
Driver for ODBC, CLI and .Net. In this case however new DB2 version 9.7
functionality is not available.

IBM Data Server drivers

The IBM data server drivers include:
򐂰 IBM Data Server Driver for JDBC and SQLJ
򐂰 IBM Data Server Driver for ODBC, CLI, and .NET

As of DB2 version 9.5 both clients and drivers are decoupled from the server
release schedule and can be downloaded separately. The IBM Data Server
Driver for JDBC and SQLJ is already included in the IBM Data server Runtime
Client. It provides support for JDBC 3 and 4 compliant applications, as well as for
Java applications using static SQL (SQLJ). Support is also provided for
pureXML, SQL/XML and XQuery. All this and other features such as connection
concentration, automatic client reroute, and more are provided within in a single
package called db2jcc4.jar. The IBM Data Server driver for ODBC, CLI and
.Net is a lightweight deployment solution for Windows applications to provide
runtime support for applications without having the need to install the Data
Server client or the Data Server Runtime Client. On Windows, the driver comes
as an installable image including merge modules to easily embed it in a Windows
installer-based installation. On Linux and UNIX there is another easy deployment
solution called the IBM Data Server Driver for ODBC and CLI, which is available
in tar format.

Communication protocols
The primary DB2 communication protocols are
򐂰 TCP, IPv4, IPv6, and Named Pipes (Windows only) for remote connections
򐂰 Interprocess Communication (IPC) for local connections within a DB2
instance

For client-server communication, DB2 supports TCP/IP and Named Pipes for
remote or local loopback connections and uses IPC for client connections, which

Chapter 1. DB2 for Linux, UNIX, and Windows 23

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

are local to the DB2 server instance. Local and remote DB2 connections are
illustrated in Figure 1-11.

Figure 1-11 Client connection scenario

Protocols are automatically detected and configured during an instance creation.

The DB2COMM registry variable identifies the protocol detected in a server. To
enable a specific protocol on the server, the db2set db2comm command must be
executed. For TCP/IP, a unique port address has to be specified in the database
manager configuration. This port is registered in the services file (usually
/etc/services on UNIX and Linux). For example, to reserve port 50000 with the
service name db2icdb2, the entry in services file is:
db2icdb2 50000/tcp

From the command line this information can be then updated in the database
manager with the following DB2 command:
db2 UPDATE DBM CFG USING SVCENAME db2icdb2

These tasks can also be performed using the DB2 Configuration Assistant utility.

Client-server configuration methods

The following methods are available for client-server configuration:
򐂰 Statically using Command Line Processor (CLP) or Configuration Assistant
(Discovery Service available)
򐂰 Dynamically using Lightweight Directory Access Protocol (LDAP)

At the client side the database information is configured using either the CATALOG
command or using the Configuration Assistant. The databases are configured
under a node, which describes host information such as protocol use, port

24 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

number, and so on. To configure a remote TCP/IP node following command is

used:
db2 CATALOG TCPIP NODE node-name REMOTE host-name SERVER service-name

The service name registered in the server or the port number can be specified in
the SERVER option. To catalog a database under this node, the command used
is:
db2 CATALOG DATABASE database-name AS alias-name AT NODE node-name

When using the Configuration Assistant GUI tool to add a database connection,
a database discovery can be started to find the desired database.

For information on how to enable and configure DB2 in an LDAP environment,

have a look at the Lightweight Directory Access Protocol (LDAP) section within
the Database fundamentals chapter in the DB2 Information Center. The DB2
Information Center contains searchable and structured Web pages and can be
installed locally. Additionally, the Information Center is also publicly available
online at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

Note: DB2 Discovery method is enabled at the instance level using the
DISCOVER_INST parameter, and at database level using DISCOVER_DB
parameter.

1.5.2 Application access

When deploying applications with DB2 Data Server 9.7 different methods can be
used:
򐂰 Single-tier
In this configuration the application and the database reside on the same
system. In enterprise environments, it may be rare to see such a
configuration, because remote access to a database server is typically
required. Nonetheless, this is quite common for developing applications that
can later be deployed transparently in a multi-tier DB2 environment without
any changes or batch applications.
򐂰 Client/Server or 2-tier
The application and the database reside on separate systems. The machines
where the application runs typically have a DB2 client installed, which
communicates over the network to a database server. For the application, the
physical location of the data is transparent. The application communicates
with the DB2 client using a standard interface (for example, ODBC) and the
DB2 client takes over the task of accessing the data over the network. In

Chapter 1. DB2 for Linux, UNIX, and Windows 25

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

some cases, such as browser or Java based access, it is not necessary to

have the DB2 client running on the same machine where the application
executes.
DB2 provides exceptional flexibility for mixing and matching client and server
platforms in a heterogeneous environment. DB2 client and server code is
available for a wide variety of platforms. For example, the application can
execute on a Windows based machine with a DB2 client for Windows, which
can then access a DB2 database on a Linux server. Likewise, the Linux
machine can act as a client and access data from UNIX servers or
mainframes.
򐂰 Multi-tier
In a multi-tier configuration, the application, DB2 client, and the data source
typically reside on separate systems. Examples of such configuration
scenarios are illustrated in Table 1-6.

Table 1-6 Multi-tier configuration examples

Client Middle-tier Server

Web-browser Web server DB2 database

Application client Application server DB2 database server 1

DB2 client DB2 database server 2

Application DB2 Connect Gateway zSeries®, iSeries®

DB2 Client

Application DB2 server Secondary Data Sources

DB2 Client (for example, Mainframe DB2,
Non-DB2, non-relational)

IBM recognizes that in many cases there may be a need for accessing data
from a variety of distributed data sources rather than one centralized
database. The data sources can be from IBM, such as DB2 or Informix, or
non-IBM databases, such as Oracle®, or even non-relational data, such as
files or spreadsheets. As illustrated in the last scenario in the Table 1-6 on
page 26, IBM offers the most comprehensive business integration solution by
allowing federated access to a variety of distributed data sources.

1.5.3 DB2 application programming interfaces (APIs)

In order to access or manage DB2 objects, several different programming
interfaces can be used as seen in Figure 1-12 on page 27.

26 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

Figure 1-12 Application connections to DB2

DB2 administrative API

DB2 provides numerous administrative APIs, which allow applications to perform
database administration tasks available in DB2 Control Center. Some of these
include import and export of data, creating, activating, backing up, or restoring a
database. These calls can be included within embedded SQL and DB2 CLI
applications. Examples of API programs can be found in the DB2 home directory
sqllib/sample/ for different programming languages. For additional information
refer to DB2 Administrative API Reference, SC23-5824.

Embedded SQL statements in applications

Two different kind of SQL statements have to be distinguished:
򐂰 Static SQL statements
With static SQL statements, SQL statement type and the table and column
names are known before compile time. Specific data values the statement is
searching for or updating are not known. These values can be represented in
host language variables.
Before compiling and linking the program, pre-compiling and binding of the
embedded SQL statements must be performed. Pre-compiling converts

Chapter 1. DB2 for Linux, UNIX, and Windows 27

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

embedded SQL statements into DB2 run-time API calls that a host compiler
can process to create a bind file. The bind command creates a package in the
database. This package then contains the SQL operation and the access plan
that DB2 will use to perform the operations.
򐂰 Dynamic SQL
Dynamic SQL statements in an application are built and executed at runtime.
For a dynamically prepared SQL statement, the syntax has to be checked and
an access plan has to be generated during the program execution.
Examples of embedded static and dynamic SQL can be found in the DB2
home directory: sqllib/samples/.

DB2 Call Level Interface (DB2 CLI)

DB2 CLI is a programming interface that can be used from C and C++
applications to access DB2 data servers. DB2 CLI is based on the Microsoft®
Open Database Connectivity (ODBC) specification, and the ISO CLI standard.
The DB2 CLI library can be loaded as an ODBC driver by an ODBC driver
manager. DB2 CLI includes support for many ODBC and ISO SQL/CLI functions,
as well as DB2 specific functions. Figure 1-13 on page 28 illustrates the ODBC
driver manger and DB2 CLI Environment.

Figure 1-13 ODBC/CLI driver connectivity

When using DB2 CLI, the application passes dynamic SQL statements as
function arguments to the database manager for processing. Because of this,

28 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

applications use common access packages provided with DB2. DB2 CLI
applications do not need to be pre-compiled or bound. Only compiling and linking
of the application is needed. Before DB2 CLI or ODBC applications can access
DB2 databases, the DB2 CLI binds files, which come with the IBM Data Server
Client to each DB2 database that will be accessed. This occurs automatically
with the execution of the first statement.

Typically, when building an ODBC application, an ODBC driver manager is

needed, which is provided by platform vendors like Microsoft and others. An
ODBC driver manager is available for Linux at:
http://www.unixodbc.org/

In environments without an ODBC driver manager, DB2 CLI is a self sufficient

driver, which supports a subset of the functions provided by the ODBC driver.
Examples of C programs using CLI calls can be found in the DB2 home directory:
sqllib/samples/cli For additional information regarding CLI, refer to Call Level
Interface Guide and Reference, Volume 1 and Volume 2, SC23-5844 and
SC23-584.

Java database connectivity application (JDBC)

DB2 Java support includes JDBC, a vendor-neutral dynamic SQL interface that
provides data access to the application through standardized Java methods. For
detailed information on the Java support, we strongly recommend the following
articles and publications:
http://www.ibm.com/support/docview.wss?uid=pub1sc23585301
http://www.ibm.com/developerworks/db2/library/techarticle/dm-0804zikopoulos/

Similar to DB2 CLI, pre-compile or bind a JDBC program is not necessary. As a

vendor-neutral standard, JDBC applications offer increased portability. The
JDBC API, which is similar to the CLI/ODBC API, provides a standard way to
access databases from Java code. The Java code passes SQL statements to the
DB2 JDBC driver, which handles the JDBC API calls. Java's portability enables
the delivery of DB2 access to clients on multiple platforms, requiring only a
Java-enabled Web browser, or a Java runtime environment™.

DB2 Data Server 9 offers different ways of creating Java applications, either
using a type 2 or type 4 JDBC driver:
򐂰 Type 2 driver:
With a type 2 driver, calls to the JDBC application driver are translated to Java
native methods. The Java applications that use this driver must run on a DB2
client, through which JDBC requests flow to the DB2 server.

Chapter 1. DB2 for Linux, UNIX, and Windows 29

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

Tip: If prototyping CLI calls before placing them in a program, use the
db2cli.exe (Windows) or db2cli (Linux) file in the sqllib/samples/cli
directory.

Note: The Type 2 driver is still included, but was deprecated. It is

recommended to convert any existing applications to use the Type 4 driver.

Figure 1-14 illustrates the Java type 2 driver connectivity.

Figure 1-14 Java type 2 driver connectivity

򐂰 Type 4 driver:
The JDBC type 4 driver can be used to create both Java applications and
applets. To run an applet that is based on the type 4 driver, a Java enabled
browser is required, which downloads the applet and the JDBC driver
(db2jcc4.jar). To run a DB2 application with a type 4 driver, an entry for the
JDBC driver in the class path and no DB2 client is required.
The JDBC driver is included in the IBM Data Server Client Driver for JDBC
and SQLJ and is architected as an abstract JDBC processor that is

30 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

independent of driver-type connectivity or target platform. Examples of JDBC

calls can be found in sqllib/samples/java/jdbc.
Figure 1-15 on page 31 illustrates the java type 4 driver connectivity
environment.

Figure 1-15 Java type 1 driver connectivity

Embedded SQL for Java (SQLj)

DB2 Java embedded SQL (SQLj) support is provided by the IBM Data Server
Runtime Client. With DB2 SQLj support, in addition to DB2 JDBC support, SQLj
applets, applications and stored procedures can be built to contain static SQL
and use embedded SQL statements that are bound to a DB2 database.

SQLj applications use JDBC as a foundation for tasks such as connecting to

databases and handling SQL errors, but also contain embedded static SQL
statements in separate SQLj source files. Unlike any other languages that can
contain embedded SQL (COBOL, C, C++) the Java code is not precompiled,
instead the SQLj translator converts SQLj clauses into JDBC statements. As
SQLj shares its underlying connection with that of JDBC, applications, it can
connect to DB2 using either type 2 or type 4 drivers. Examples of SQLj calls can
be found in sqllib/samples/java/sqlj.

Chapter 1. DB2 for Linux, UNIX, and Windows 31

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

ActiveX Data Objects and Object Linking and Embedding

(Windows only)
DB2 supports ActiveX® Data Object (ADO) applications that use the Microsoft
OLE DB to ODBC bridge. ActiveX Data Objects (ADO) provides the ability to
write applications to access and manipulate data in a database server through
an OLE DB provider. This means applications will have uniform access to data
stored in diverse information sources.

Remote Data Objects (RDO) provides an information model for accessing

remote data sources through ODBC. RDO offers a set of objects that make it
easy to connect to a database, execute queries and stored procedures,
manipulate results, and commit changes to the server. As RDO implements a
thin code layer over the ODBC API, it requires an ODBC data source to be
created for the DB2 database that you are connecting to.

ADO.NET
DB2 supports Microsoft ADO.NET programming interface through a native
managed provider. Such applications can use the DB2 .Net, the OLE DB .Net or
ODBC .NET Data Provider. High-performing Windows Forms, Web Forms, and
Web Services can be developed using the ADO.NET API. DB2 supports a
collection of features that integrate seamlessly into Visual Studio 2003, 2005 and
2008 to make it easier to work with DB2 servers and develop DB2 procedures,
functions and objects.

The IBM Data Server Provider for .Net extends data server support for the
ADO.NET interface and delivers high-performing, secure access to IBM data
servers:
򐂰 DB2 Version 9 (or later) for Linux, UNIX, and Windows
򐂰 DB2 Universal Database™ Version 8 for Windows, UNIX, and Linux
򐂰 DB2 for z/OS and OS/390® Version 6 (or later), through DB2 Connect
򐂰 DB2 for i5/OS Version 5 (or later), through DB2 Connect
򐂰 DB2 Universal Database Version 7.3 (or later) for VSE and VM, through DB2
Connect
򐂰 IBM Informix Dynamic Server, Version 11.10 or later
򐂰 IBM UniData®, Version 7.1.11 or later
򐂰 IBM UniVerse, Version 10.2 or later

When used in conjunction with stored procedures and the federated database
capabilities of DB2 data servers and DB2 Connect servers, this data access can
be extended to include a wide variety of other data sources, including non-DB2
mainframe data, Informix Dynamic Server (IDS), Microsoft SQL Server®, Sybase

32 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch01.fm

and Oracle databases as well as any data source that has an OLE DB Provider
available.

For more information about developing ADO.NET and OLE DB refer to DB2 for
Linux, UNIX, and Windows Developing ADO.NET and OLE DB Applications,
SC23-5851-01 available at:
http://www.ibm.com/support/docview.wss?uid=pub1sc23585101

Perl DBI
DB2 supports the Perl Database Interface (DBI) specification for data access
through the DBD::DB2 driver. The Perl DBI module uses an interface that is
similar to the CLI and JDBC interfaces, which makes it easy to port Perl
prototypes to CLI and JDBC. As of DB2 9.5 the Perl DBI comes with support for
DB2 pureXML technology. This allows you to insert XML documents without the
need to parse or validate XML. The Perl driver also supports multi-byte character
sets, which means your application does not have to deal with the conversion
itself when interacting with the database.

More information according Perl DBI can be found at:

http://www.ibm.com/software/data/db2/perl/

PHP
The “PHP: Hypertext Preprocessor” is a modular and interpreted programming
language intended for the development of Web applications. Its functionality can
be customized through the use of extensions. DB2 supports PHP through an
extension called pdo_ibm, which allows DB2 access through the standard PHP
Application Objects (PDO) interface. In addition the ibm_db2 extension offers a
procedural API that, in addition to the normal creates, read, update, and write
database operations, also offers extensive access to the database metadata.
The most up to date versions of ibm_db2 and pdo_ibm are available from the
PHP Extension Community Library (PECL):
http://pecl.php.net/

For DB2 data server for Linux, UNIX and Windows pre-compiled and easy to
install packages called “Zend Core for IBM” are available for download. Zend
Core for IBM features tight integration with DB2 and Informix Dynamic Server
drivers. This can be used in a development or production system. More
information about Zend Core for IBM can be found at:
http://www.ibm.com/software/data/info/zendcore/?S_TACT=appdmain&S_CMP=ibm_im

Ruby on Rails
Since DB2 9.5 has drivers for Ruby, which is an object-oriented programming
language. Combined with the open source Ruby framework called Rails the

Chapter 1. DB2 for Linux, UNIX, and Windows 33

7093ch01.fm Draft Document for Review October 16, 2009 10:21 am

development of web-based and database-driven applications can be extremely

quick. Included in DB2 is the IBM_DB Ruby adapter, which allows any
database-backed Ruby application to interface with IBM data servers.

For more information about IBM Ruby projects and the RubyForge open source
community, refer to the following website:
http://rubyforge.org/projects/rubyibm/

34 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

Chapter 2. MySQL database

When planning a conversion project it is critical to understand how the source
data server operates, in order to successfully convert all functionality to the
destination data server. The goal of this chapter is to discuss the MySQL
database architecture and design, while keeping in mind how the features will be
converted to DB2.

The following MySQL details are discussed:

򐂰 License and origin
򐂰 Architecture overview
򐂰 Data handling
򐂰 Utilities

© Copyright IBM Corp. 2009. All rights reserved. 35

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

2.1 MySQL licensing overview

MySQL is a dual licensed relational database management system developed
and distributed by MySQL AB (http://www.mysql.com), a company, which builds its
business by providing services around MySQL. In late 2007 MySQL AB was
acquired by Sun™ Microsystems, which was later acquired by Oracle in April of
2009. At the time of writing this book, the acquisition of Sun Microsystems by
Oracle is still ongoing and the future of MySQL licensing model is still uncertain.
Currently MySQL implements a dual licensing model. The first is a General
Public License (GPL), the “open source” license, which is mostly linked to open
source projects. That is, if you implement this licensing model and distribute
MySQL with applications, the application source code must be open source as
well. The second licensing model is for commercial use covering all other
environments outside of the purely open source and non-commercial
environments. The commercial license has a significant price tag equal to the
prices of other commercial database products.

DB2 offers a no-charge community edition (DB2 Express-C) of the DB2 data
server. This edition of DB2 is completely free to develop, deploy, and to
distribute. DB2 Express-C is a full-function relational and XML data server and
has the same reliability, flexibility, and power of the higher editions of DB2. DB2
also offers the DB2 Express + Fix Term license option, which is price comparable
with the MySQL Enterprise Gold pricing. For more details on each of the DB2
edition available, refer to 1.2.1, “DB2 Data Server Editions for the production
environment” on page 3.

MySQL was initially developed for UNIX and Linux applications. It became
popular when Internet Service Providers (ISP) discovered that MySQL could be
offered free of charge to their Internet customers providing all of the storage and
retrieval functionality that a dynamic Web application would need. It was also
advantageous since ISPs primarily use Linux or UNIX as their base operating
system, in combination with APACHE as their favorite Web server environment.
Today MySQL is also used as an integrated or embedded database for various
applications running on almost all platform architecture.

2.2 MySQL architecture overview

MySQL is built on a client-server architecture model that can be installed on
Windows, UNIX, Macintosh, and Linux platforms. The server is the host that
manages the database objects and data, whereas the client requests and works
with the actual data. MySQL client programs connect to the MySQL server either
through a local or remote connection. If the client is running remotely, MySQL

36 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

always connects using TCP/IP. If the client is running on the local server any of
the supported connection protocols can be used.

Figure 2-1 illustrates the conceptual architecture of the MySQL database. The
next several sections cover the functionality of the integrated components in
more detail.

Figure 2-1 Conceptual MySQL architecture

2.2.1 Database client and non-client utilities

The client layer represents the interface between the user and the database.
Most components are provided by MySQL AB in the server installation bundle.
The graphical tools need to be downloaded and installed separately. MySQL
clients include the:
򐂰 Query interface
򐂰 Administrative interface and utilities
򐂰 Application interface and utilities

The client layer is the front end component with which users will interact. This
component presents three types of users that interact with the server: query
users, administrators, and applications.

Chapter 2. MySQL database 37

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

Query users
Query users can interact with the database server through a query interface
called mysql, which allows users to issue SQL statements and view the results
returned from the server using the command line. There is also a graphical tool
called MySQL Query Browser that provides a graphical interface to create and
execute database queries.

In DB2 you can use the command line process for the same functionality. Use
the db2 or db2cmd command to start the command line processor.

Administrators
Administrators use the administrative interface and utilities, such as mysqladmin
and MySQL Administrator. Theses tools can be used for creating or dropping
databases and users as well as managing the MySQL server. These tools
connect to the database server using the native C client library. There are also
utilities that can be used for administrative purposes but do not connect to the
MySQL server, instead work directly with the database files.These tools are
myisamchk for table analysis, optimization, and crash recovery and myisampack
for creating read-only compressed versions of MyISAM tables.

DB2 offers a rich set of database management GUI tools, such as the DB2
Control Center, the Optim Database Administrator, and the IBM Optim Data
studio. These tools simplify database administration by providing one single tool
to completely manage your entire database environment. You can use these
tools to also query the database. The GUI tools are discussed in detail in 9.8,
“Database management tools” on page 312.

Applications
Applications communicate with the database server through MySQL APIs
available for various programming languages such as C++, PHP, Java, Perl,
.NET, and so on. We discuss this in more detail later in the chapter.

2.2.2 Database server

The database server represents the core functionality of the database
architecture. It is responsible for storing, accessing, and managing the
databases and database objects. MySQL server is multi-threaded allowing
multiple users to access multiple databases. Figure 2-1 on page 37, shows the
different components of the MySQL server. As shown in the figure, you can find
the following main components in MySQL:
򐂰 Connection pool
򐂰 SQL interface
򐂰 Parser

38 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

򐂰 Optimizer
򐂰 Caches and buffers
򐂰 Database management utilities
򐂰 Storage engines
򐂰 Physical resources

Connection pool
The connection pool assigns user connections to the database and
corresponding memory caches. The utilities and programs that are included with
MySQL connect using the Native C API. Other applications can connect using
the standard drivers MySQL offers, such as C++, PHP, Java, Perl, .NET, and so
on. MySQL supports TCP/IP, UNIX socket file, named pipe and shared memory
networking protocols depending on the type of operating system used. For more
details on application programming interfaces, see 2.5, “MySQL application
programming interfaces (API)” on page 51.

SQL interface
The SQL interface accepts and conveys the SQL statements from the
connecting user or MySQL application. This layer is independent of the storage
ending layer and therefore SQL support statements are not dependent on the
type of storage engine being used. The SQL statement is then passed to the
SQL parser for further processing.

SQL parser
The parser analyses the SQL statements and validates the SQL query syntax.
The parser breaks up the statement and creates a parse tree structure to
validate the SQL query syntax and prepare the statement for the optimizer.

SQL optimizer
The SQL optimizer verifies that the tables exist and that the records can be
access by the requesting user. After security checking, the query is analyzed and
optimized to raise performance of the query process.

Cache and buffers

Caches and buffers are used to increase the speed of the server by decreasing
the amount of data that the MySQL server needs to read from disk. The memory
caches are used for both global and engine specific caches. MySQL uses a
query cache to increase performance of data queries. The query cache holds
both the query and the result of the query. There is also a cache to store the
table descriptors and a cache to hold the host name. MySQL stores the user
account information in the grant table buffers and the MyISAM index blocks in
the key buffer.

Chapter 2. MySQL database 39

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

Database management utilities

The Database management utilities assist users and DBAs with managing and
administrating database objects. The majority of these tools are independent of
the storage system being used. The following are the main components of
Database management utilities:
򐂰 Administration
Administrative utilities are used to help manage the database server,
database objects, and data. Some of the utilities that are available with
MySQL can assist with compressing MyISAM tables, managing MyISAM log
files, managing and repairing tables, checking user privileges, and converting
tables' storage engine.
򐂰 Backup and recovery
Backups can be used for restoring a system crash, hardware update or for
replicating a test or development environment. MySQL has multiple
techniques to backup a database. If a SQL level backup is needed, the
SELECT INTO command can be used to output the table data to a file. The
mysqlhotcopy script can be used to make a fast copy of a database or an
individual table. The mysqldump utility can be used to backup a single or
multiple databases.
򐂰 Replication
Replication allows for a secondary server, called the slave, to maintain a copy
of the master server's database without having to perform a full backup and
recovery. The slave database is updated with only the recent changes that
occurred on the master. There can be one or more slave systems.

The following are additional components of database management utilities.

򐂰 Security
򐂰 Configuration
򐂰 Conversion and metadata

Pluggable storage engine

The MySQL storage engine is responsible for accessing and modifying the
stored data. MySQL supports multiple storage engines, each with its own
characteristics. The DBA can choose a specific storage engine depending on
how the data is going to be accessed and used. Storage engines are allocated
on a table basis. It is possible to mix and match different storage engines within a
database. If no storage engine is specified the default MyISAM storage engine
will be used. We discuss each storage engine in section 2.3.2, “MySQL storage
engines” on page 44.

40 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

Physical resource
This is the bottom layer of the MySQL architecture, and represents the
secondary storage or physical disk. This layer is accessed through the storage
engines to store or retrieve data. The actual data stored in this layer is:
򐂰 Data files (user data)
򐂰 Data dictionary (metadata)
򐂰 Indexes
򐂰 Log information
򐂰 Statistical data

In the next section, we cover how the database objects and data are physically
stored on the server.

2.3 MySQL design and SQL compliance

In this section, the following topics are discussed:
򐂰 MySQL directory structure
򐂰 MySQL storage engines
򐂰 MySQL standard SQL compliance

2.3.1 MySQL directory structure

MySQL can be installed as one or more databases within one server. For each
database, a directory is created with the same name as the database. This
directory holds all of the database data and indexes. On Linux distributions,
when installing using RPM packages, the default directory for the database and
log files is /var/lib/mysql. Installation of MySQL can also be performed using
source distribution, where the default directory and location of all MySQL files
can be chosen. For UNIX distributions, the databases are stored under
/usr/local/mysql/data. For Windows this is under
C:\Program Files\MySQL\MySQL Server 5.1\data. Table 2-1 shows the default
installation directories for MySQL Windows, Linux, and UNIX installations.

Table 2-1 MySQL default directories

Contents Windows Default Linux Default UNIX Default
Directory Directory Directory

Default C:\Program Files\ /usr/local/mysql

installation MySQL\MySQL Server
directory 5.1

Chapter 2. MySQL database 41

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

Databases <Default Dir>\data /var/lib/mysql <Default Dir>/data

and log files

Client <Default Dir>\bin /usr/bin <Default Dir>/bin

programs
and scripts

mysqld <Default Dir>\bin /usr/sbin <Default Dir>/bin

server

Manual <Default Dir>\Docs /usr/share/info <Default Dir>/docs

Example <Default Dir>\examples

scripts

Include files <Default Dir>\include /usr/include/mysq <Default Dir>/include

l

Libraries <Default Dir>\lib /usr/lib/mysql <Default Dir>/lib

Utilities and <Default Dir>\scripts <Default Dir>/scripts

scripts

Error <Default Dir>\share /usr/share/mysql <Default

message Dir>/share/mysql
files

In the example in Figure 2-2 on page 42 there are two databases on this MySQL
server. The first database is the mysql database, which by default holds the
security information. The second database is the sample database inventory,
which is discussed in more detail in Chapter 4, “Conversion scenario” on
page 75.

Figure 2-2 MySQL example directory structure

42 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

The following files are created for each database directory in the MySQL home
directory:
򐂰 Files with the frm extension contain the structure definition of table and view,
known as the schema.
򐂰 Files with the MYD extension contains the table data.
򐂰 Files with the MYI extension contains the table indexes.
򐂰 If there are triggers, then there also are files with the TRN and the TRG
extension.

Example 2-1 shows the files created for each table in our sample database.

Example 2-1 MySQL example table files

mysqlServer:/var/lib/mysql/inventory # ls -lrt
total 428
-rw-rw---- 1 mysql users 8670 Jul 21 23:11 locations.frm
-rw-rw---- 1 mysql users 65 Jul 21 23:11 db.opt
-rw-rw---- 1 mysql users 4976 Jul 21 23:15 locations.MYD
-rw-rw---- 1 mysql users 4096 Jul 21 23:18 locations.MYI
-rw-rw---- 1 mysql users 8918 Aug 7 13:21 owners.frm
-rw-rw---- 1 mysql users 8804 Aug 7 13:29 inventory.frm
-rw-rw---- 1 mysql users 1584 Aug 7 14:21 empGroup.frm
-rw-rw---- 1 mysql users 1585 Aug 7 14:22 techGroup.frm
-rw-rw---- 1 mysql users 1587 Aug 7 14:23 bossGroup.frm
-rw-rw---- 1 mysql users 8746 Aug 7 17:23 groups.frm
-rw-rw---- 1 mysql users 1595 Aug 10 01:42 managerGroup.frm
-rw-rw---- 1 mysql users 1596 Aug 10 01:43 generalGroup.frm
-rw-rw---- 1 mysql users 1587 Aug 10 02:58 testGroup.frm
-rw-rw---- 1 mysql users 128 Aug 10 02:58 groups.MYD
-rw-rw---- 1 mysql users 8886 Aug 10 03:29 services.frm
-rw-rw---- 1 mysql users 8620 Aug 10 03:54 status.frm
-rw-rw---- 1 mysql users 2048 Aug 10 03:54 status.MYI
-rw-rw---- 1 mysql users 160 Aug 10 03:54 status.MYD
-rw-rw---- 1 mysql users 8692 Aug 10 03:56 severity.frm
-rw-rw---- 1 mysql users 2048 Aug 10 15:31 groups.MYI
-rw-rw---- 1 mysql users 37876 Aug 11 15:45 inventory.MYD
-rw-rw---- 1 mysql users 27648 Aug 11 16:39 inventory.MYI
-rw-rw---- 1 mysql users 212 Aug 24 16:07 severity.MYD
-rw-rw---- 1 mysql users 40 Aug 24 16:08 updateDate.TRN
-rw-rw---- 1 mysql users 378 Aug 24 16:08 services.TRG
-rw-rw---- 1 mysql users 2048 Aug 27 20:21 severity.MYI
-rw-rw---- 1 mysql users 45576 Sep 9 03:08 services.MYD
-rw-rw---- 1 mysql users 37888 Sep 11 00:46 services.MYI
-rw-rw---- 1 mysql users 30720 Sep 11 13:41 owners.MYI
-rw-rw---- 1 mysql users 61144 Sep 11 13:41 owners.MYD

Chapter 2. MySQL database 43

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

Log files by default are created in the mysql home directory. The security data
tables in mysql database are in the directory /< mysql home directory>/mysql.
Table 2-2 lists the files of the security data tables.

Table 2-2 Permission information stored in tables

File Description

columns_priv.MYD, Permission for individual columns within a table

columns_priv.MYI,
columns_priv.frm

db.MYD, Permission for individual database

db.MY,
db.frm

func.MYD, Permission for user defined functions

func.MYI,
func.frm

host.MYD, General permission by host

host.MYI,
host.frm

procs_priv.MYD, Permission for stored procedures

procs_priv.MYI,
procs_priv.frm

tables_priv.MYD, Permission for individual tables within a database

tables_priv.MYI,
tables_priv.frm

user.MYD, General Permission by user

user.MYI,
user.frm

By default DB2 uses a better approach for the logical and physical distribution of
database objects. Different from MySQL, DB2 stores all database objects in
table spaces. The benefits of table spaces are increased performance and
simplified management. To take advantage of this advanced database
distribution, refer to 6.2.1, “Database manipulation” on page 123, where we
discuss in detail the conversion of a MySQL database structure to DB2.

2.3.2 MySQL storage engines

Although, MySQL supports multiple storage engines per database, each table
must be managed by a specific storage engine. All storage engines will create

44 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

table definitions in a file with the .frm extension upon creation. MySQL storage
engines can be split into two different categories:
򐂰 Non-transaction-safe
򐂰 Transaction-safe

Transaction-safe storage engines have commit and rollback capabilities and can
be recovered from a crash. Therefore these storage engines guarantee the
Atomicity, Consistency, Isolation, and Durability (ACID properties) of a database.
Non-transaction-safe storage engines are faster and require less memory and
disk space. However, non-transaction-safe storage engines do not guarantee the
database is left in a consistent state, as they do not support ACID properties

The following are the storage engines, each storage engine are explained in this
section.
򐂰 Non-transaction-safe tables can be managed by the following storage
engines:
– MyISAM
– Memory
– Merge
– Archive
– CSV
– Federated
– Blackhole
򐂰 Transaction-safe tables can be managed by the following storage engines:
– InnoDB

MySQL supports transactions with the InnoDB and NDB transactional storage
engines. Though both provide full ACID compliance, the performance and
throughput may be a concern and it is often a reason to convert to DB2.

In DB2 all tables support transactions. Therefore, tables that are managed by
MySQL InnonDB, MyISAM, ARCHIVE, and CSV storage engines can all be
converted to a DB2 regular table. Details on converting MySQL tables to a DB2
table are discussed in 6.2.2, “Table manipulation” on page 128.

MyISAM Storage Engine

The MyISAM storage engine, based on ISAM code, was enhanced to overcome
the disadvantages of the ISAM storage engine and provide more useful
extensions. MyISAM is the default storage engine of MySQL since version 3.23.
This storage engine stores the table definition in a file with the .frm extension, the
index is stored in a file with the .MYI extension, and the data is stored in a file
with the .MYD extension. MyISAM storage engine uses table-level locking.

Chapter 2. MySQL database 45

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

MyISAM supports three different storage formats:

򐂰 Static tables
Have a fixed length and are the default format MyISAM uses, if no
VARCHAR, VARBINARY, BLOB, or TEXT columns are used.
򐂰 Dynamic tables
Are used as default if VARCHAR, VARBINARY, BLOB, or TEXT columns are
defined in the table.
򐂰 Compressed tables
Are read-only static and dynamic tables that have been compressed. To write
data to a compressed table, the table has to be uncompressed first.

Memory storage engine

The memory storage engine, formally know as heap table, uses hashed and
B-tree indexes and can only be held in memory. Therefore, they are primarily
used as temporary tables. This storage engine only creates the .frm file for the
table definition. The Memory storage engine uses table-level locking.

Merge storage engine

The merge storage engine allows the grouping of multiple MyISAM tables across
the same disk or different disks into one table. The merge storage engine can
only be used with MyISAM tables that have the same columns and indexes. The
.MRG file contains the list of MyISAM tables that make up the new Merge table.
If a table created with this storage engine is dropped, the Merge .frm and .MRG
files are removed and the individual MyISAM tables still exist. Tables created
with the merge storage engine support SELECT, UPDATE, INSERT and
DELETE operations if the user has the privileges on the underlining MyISAM
tables.

Archive storage engine

The archive storage engine is usually used for storing large amounts of data that
will not be modified. The archive engine only supports SELECT and INSERT
operations. Data stored in the archive engine is stored in compressed format
without indexing on the data. This storage engine will store the table definition in
a file with the .frm extension, the data is stored in a file with the .ARZ extension,
and the metadata is stored in a file with the .ARM extension. The archive storage
engine uses row-level locking.

CSV storage engine

The CSV (comma separated value) storage engine stores data in comma
septerated text files. The table definition is stored on the server in a file with the
.frm extension and the data is stored in a file with the .CSV extension.

46 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

Federated storage engine

The federated storage engine allows tables on a remote server to be accessed
as if stored locally. Table definition, the .frm file, is stored locally but no data is
actually stored on the local server.

Blackhole storage engine

The blackhole storage engine creates a table where no data is stored and no
data can be retrieved. This storage engine only creates the .frm file for the table
definition.

InnoDB Storage Engine

Since MySQL 4.0 InnoDB is enabled by default. The InnoDB engine is a
complete separate database back-end produced by a company called Innobase
Oy (http://www.innodb.com) and placed under MySQL. InnoDB tables are
transaction safe and support foreign key constraints. This storage engine will
store the table definition in a file with the .frm extension. Table data and indexes
are stored in a table space consisting of several files or even raw disk partitions.
InnoDB uses B-tree indexes to locate data in the tables. This storage engine
uses row-level locking and does not lock rows for read statements. DB2 9.7 has
introduced a similar approach to locking, discussed further in 8.3.3, “DB2
isolation levels” on page 273. This changing in DB2 locking behaviour makes
porting an application much easier.

Innodb was acquired by Oracle in 2006 and this resulted into a development
effort by MySQL AB to build their own transactional safe storage engine which is
still under development at the time the book is written.

NDB cluster storage engine

The NDBCLUSTER, also known as NDB, is a transaction safe storage engine.
This storage engine was introduced in MySQL 4.1. This storage engine allows
for clustering of a database in a shared-nothing architecture. This type of
architecture allows a table to be split apart across multiple servers, where each
server has its own set of data. In order for the table to be a part of the cluster, the
table must be using NDBCLUSTER storage engine. As of MySQL 5.1.24 support
for NDBCLUSTER storage engine has been removed from the standard MySQL
release and now available in its own release, known as MySQL Cluster NDB.

2.3.3 MySQL standard SQL compliance

Up to version 5.1 MySQL meets most closely to SQL-92 and meets ODBC levels
0-3.51 standards. MySQL aims towards fully complying with the ANSI/ISO SQL
standard. The most important missing compatibilities of MySQL default table
MyISAM toward the ANSI-SQL-92 standard refers to:

Chapter 2. MySQL database 47

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 Transactions
The MySQL default storage engine MyISAM does not support transactions.
For this storage engine the developers of MySQL followed the “atomic
operations” data integrity model. Auto-commit is always enabled by default for
MyISAM, therefore every time a statement is executed the changes are
committed to the database, as shown in Example 2-2.

Example 2-2 Atomic operation

INSERT INTO table1 ...
COMMIT
DELETE FROM table2 ...
COMMIT

MySQL supports transactions with the InnoDB and NDB transactional storage
engines. Both engines provide full ACID compliance. In contrast, all tables in
DB2 support transactions and provide full ACID compliance.
򐂰 Referential integrity
Referential integrity is the state in which all values of all foreign keys are valid.
The relationship between some rows of the DEPT and EMP tables, shown in
Figure 2-3 on page 48, illustrates referential integrity concepts and
terminology. For example, referential integrity ensures that every foreign key
value in the DEPT column of the EMP table matches a primary key value in
the DEPTNO column of the DEPT table.

Figure 2-3 Referential Integrity

48 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

In MySQL Server, MyISAM tables do not support foreign key constraints.

MySQL only parses the FOREIGN KEY syntax in CREATE TABLE
commands, but does not use or store this information. Only InnoDB tables
support checking of foreign key constraints including ON DELETE CASCADE
and ON UPDATE CASCADE.
DB2 supports referential integrity on all tables created. In 6.5, “Sample
database conversion” on page 148 we discuss how to take advantage of
foreign keys in DB2.

2.4 MySQL utilities

MySQL is distributed with a set of support utilities. The Internet however provides
a larger set of third party tools to manage MySQL databases. In this section, we
attempt to give a brief overview of the MySQL distributed set of supported tools.
For more information visit:
http://dev.mysql.com/doc/refman/5.1/en/programs.html

DB2 has a full set of utilities available to work with and manage the database
environment. DB2 utilities are described in 1.4, “DB2 utilities” on page 18.

2.4.1 Overview of the MySQL server-side programs and utilities

This section introduces the distributed set of server-side tools included in the
MySQL package.

MySQL server programs

These programs are used to start, stop, and manage the MySQL server for
different purposes:
򐂰 mysqld
򐂰 mysqld-nt
򐂰 mysqld_safe
򐂰 mysqld-debug
򐂰 mysqld_multi
򐂰 mysql.server
򐂰 mysqld-max
򐂰 mysqlmanager

Setup programs
The rest of the programs are used for setting up operations during the installation
or upgrade of the MySQL server:

Chapter 2. MySQL database 49

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 mysql_install_db
򐂰 mysql_fix_privilege_tables
򐂰 make_binary_distribution
򐂰 mysqlbug
򐂰 comp_err
򐂰 make_win_bin_dist
򐂰 mysql_secure_installation
򐂰 mysql_tzinfo_to_sql
򐂰 mysql_upgrade

2.4.2 Overview of the MySQL client-side programs and utilities

The distributed set of client-side tools provided by MySQL AB are:
򐂰 mysql
򐂰 mysqladmin
򐂰 mysqlcheck
򐂰 mysqldump
򐂰 mysqlimport
򐂰 mysqlshow
򐂰 mysqlaccess
򐂰 mysqlbinlog
򐂰 mysqlhotcopy
򐂰 myisamchk
򐂰 myisampack
򐂰 innochecksum
򐂰 my_print_defaults
򐂰 myisamlog
򐂰 mysql_convert_table_format
򐂰 mysql_fix_extensions
򐂰 mysql_tableinfo
򐂰 mysql_setpermissions
򐂰 mysql_waitpid
򐂰 mysql_zap
򐂰 replace
򐂰 perror
򐂰 msql2mysql
򐂰 mysql_config
򐂰 MySQL Administrator
򐂰 MySQL Query Browser

50 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

2.5 MySQL application programming interfaces (API)

The MySQL API is an interface by which an application program communicates
with the MySQL database. This section gives an overview of which APIs are
available for MySQL.

Generally speaking there are four main approaches to connect to a MySQL

database as shown in Figure 2-4 on page 51.
򐂰 JDBC with Connector/J
򐂰 .NET with Connector/NET
򐂰 ODBC with Connector/ODBC
򐂰 Other APIs with C Library

Figure 2-4 MySQL APIs

The first approach is to connect the Java application using JDBC and the
Connector/J, which is officially supported by MySQL. This connector is written in
Java and does not use the C client library to implement the client/server
communication protocol.

The second approach is to connect the .NET application using Connector/NET

which is written in C# and does not use the C client library to implement the
client/server communication protocol. This approach is officially supported by
MySQL AB.

The third approach is to connect using the Connector/ODBC for applications that
use ODBC standards. This connector uses the embedded C client libraries to
implement the client/server communication protocol. This approach is officially
supported by MySQL.

Chapter 2. MySQL database 51

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

The fourth approach is to use the third-party APIs provided by the programming
languages like PHP, Perl, or Python. These APIs will use the embedded C client
libraries to implement the client/server communication protocol. The third-party
APIs are not officially supported by MySQL. The following lists some of the APIs
available for MySQL:
򐂰 C API
The API to connect from C programs to a MySQL database. More details can
be found at:
http://dev.mysql.com/doc/refman/5.1/en/c.html
򐂰 C++ API
The API to connect to a MySQL database from C++. Information can be
found at:
http://forge.mysql.com/wiki/Connector_C%2B%2B_Binary_Builds
򐂰 PHP API
PHP contains support for accessing several databases including MySQL.
Information about MySQL access can be found in the PHP documentation,
which can be downloaded at:
http://www.php.net/download-docs.php
򐂰 PERL API
The Perl API consists of a generic Perl interface and a special database
driver. The generic interface in Perl is called Database Interface (DBI) and for
MySQL the driver is called DBD::mysql. Information about the DBI can be
found at:
http://dbi.perl.org/
򐂰 Python API
The API to connect to MySQL for Python is called MySQLdb, and can be
found at:
http://sourceforge.net/projects/mysql-python/
򐂰 Ruby API
The API for accessing MySQL servers from Ruby programs and can be found
at:
http://tmtm.org/en/mysql/ruby/

DB2 supports the most frequently used MySQL programming languages. These
languages include PHP, Java, Perl, Python, Ruby, C#, C/C++, and Visual
Basic®. With proper planning and knowledge these applications can be
converted to DB2 with minimal effort. In Chapter 8, “Application conversion” on

52 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch02.fm

page 207 we discuss and provide examples for converting applications from
MySQL to DB2.

Chapter 2. MySQL database 53

7093ch02.fm Draft Document for Review October 16, 2009 10:22 am

54 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

Chapter 3. Planning the conversion

from MySQL to DB2
Proper planning influences the success of a project significantly; this chapter
discusses the overall conversion planning stages. This includes the
considerations required prior to porting, various assessments that must be
considered, and finally the porting steps. In this chapter we discuss the following
conversion topics in detail:
򐂰 Project planning
򐂰 Application assessment and system planning
򐂰 The conversion process
– Porting preparation and installation
– Database structure porting
– Data porting
– Application porting
– Basic administration
– Testing and tuning

We also provide information about how IBM conversion specialists can support
you in your conversion project in any of the steps involved and the available
conversion tools such as the IBM Data Movement Tool.

© Copyright IBM Corp. 2009. All rights reserved. 55

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

3.1 Conversion project planning overview

In order to understand what needs to be done and how long it will take, each
conversion or conversion project begins with a conversion assessment. A
systematic and organized analysis provides a detailed picture of the full project.
The assessment requires good knowledge of the application to be converted and
the products that will be used.

To assess an application for conversion, an application profile that describes the

application architecture, technologies used, application functions, application
interface, application environment characteristics, database detail, application
size, and so on should be created.

Based on the application profile, you can plan software and hardware needs for
the target system. The planning stage is also a good time to consider many of the
rich functions and features of the DB2 product family, which may increase
productivity and reduce maintenance costs. While the most common reason for
one-to-one conversion is the usage of the more advance features that MySQL
does not provide, reduced runtime of transactions by multiple times also helps.
The advanced optimizer in DB2 showed improvements of approximately 20x for
real production load in various projects.

Knowledge regarding the skills required and resources available for the
conversion project is required. IBM provides a variety of DB2 courses to help
IBM customers learn DB2 quickly. For more information go to:
http://www.ibm.com/developerworks/
http://www.ibm.com/industries/education/index.jsp

and
http://www.ibm.com/developerworks/data/bootcamps

A conversion assessment provides you with the overall picture of the conversion
tasks and efforts needed. From a conversion assessment a conversion project
plan can be created to manage each conversion step.

To execute typical conversion steps, various tools are available to help you save
time in your conversion project. IBM offers the free conversion tool, the IBM
Data Movement Tool, for converting from various relational database systems to
DB2.

The process of database and application conversion consists of the following

main steps, which are discussed in detail later:
򐂰 Porting preparation and installation database structure porting
򐂰 Data porting

56 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

򐂰 Application porting
򐂰 Basic administration
򐂰 Testing and tuning
򐂰 User education

Experienced IBM specialists can support you during any phase of the conversion
project with special conversion offerings provided by IBM Worldwide.

3.1.1 Benefits of converting to DB2

DB2 offers open, industrial-strength database management for e-business,
business intelligence, transaction processing, and a broad range of applications.
At the same time a lot of focus and investment has been put into simplifying
administration within DB2 with autonomic computing.

Some primary motivators for conversion are:

򐂰 Multi platform support
DB2 is a true cross-platform database management system (DBMS), running
on a wide variety of systems including Linux, Windows 98/NT/2000/XP,
Solaris, HP-UX, and AIX. For more information about platform support see:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp
򐂰 Performance Optimization
DB2 has dominated key performance benchmarks many times over the years,
including both On-Line Transaction Processing (OLTP) and On-Line
Analytical Processing (OLAP) benchmarks. What this world class
performance means for businesses running DB is that they can do more work
with their existing hardware, thereby avoiding or delaying costly server
upgrades.
http://www.ibm.com/software/data/db2/9/editions_features_perf_ent.html
򐂰 Reliability
For many businesses, database downtime causes both money and
opportunity loss. DB2 can minimize the downtime associated with many
planned activities, such as altering a table, and many unplanned events, such
a power outage, by either eliminating the downtime completely or reducing it's
duration to a few seconds. DB2 High Availability Feature provides 24 x 7
availability for your DB2 data server through replicated failover support and
data recovery modules. For more details visit:
http://www.ibm.com/software/data/db2/9/editions_features_ha.html
򐂰 Industrial scalability

Chapter 3. Planning the conversion from MySQL to DB2 57

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

DB2 has the best scalability by being the only one to have transaction
processing and business intelligence scaling up to 1000 nodes. For more
information about scalability see:
http://www.ibm.com
򐂰 Integrated support for native environments
DB2 conforms to many standards including operating system support. It maps
closely onto internal resources for performance and scalability making it more
reliable and tightly integrated.
򐂰 Deep compression
Businesses with large volumes of data know how expensive storage can be.
DB2 can dramatically reduce this cost with industry leading data compression
technologies that compress rows, indexes, temporary tables, LOBs, XML, and
back up data with compression rates that can reach over 80%. This allows
DB2 to keep more data in memory, thereby avoiding performance-robbing
disk I/O. In turn causing database performance to increase considerably. For
more details on deep compression see:
http://www.ibm.com/software/data/db2/compression/
򐂰 Security
Unauthorized data access is an ever present threat that can cost businesses
considerable sums of money, their reputation, or both. DB2 offers a
comprehensive suite of security features that effectively and decisively
minimizes this threat. DB2 provides additional peace of mind with lightweight
security audit mechanisms that are used to verify that any unauthorized data
access.
http://www.ibm.com/software/data/db2/9/editions_features_advaccess.html
򐂰 pureXML
DB2 pureXML revolutionizes the management of XML data with breakthrough
native XML support provided only by DB2. It eliminates much of the work
typically involved in the management of XML data and it serves data at
unmatched speeds. If you work with XML data, then you need to know about
DB2 pureXML. Now DB2 9.7 for Linux, UNIX, and Windows opens new
opportunities to efficiently analyze XML data in data warehouses.
http://www.ibm.com/software/data/db2/xml/
򐂰 Integrated system management tools
DB2 has a number of tools for managing the database system. IBM Optim
Data studio is a rich and extensible solution to help you develop DB2
applications and manage databases. The Health Monitor and the Health
Center help to easily capture and monitor the overall health of the database.
The Replication Center is a tool used to setup and administer a replication

58 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

environment. The Configuration Assistant and the Control Center can be

used to assist in configuring and maintaining database objects. This is just a
small number of management tools available for DB2. For a list of all tools and
more information on the tools see:
http://www.ibm.com/software/data/db2imstools/products/db2-luw-tools.html
򐂰 Self-managing and resource tuning capability
DB2 also contains self-managing and resource tuning database technology,
such as the Self-Tuning Memory Manager (STMM) and automatic storage
which provides enhanced automation to configure, tune, and manage
databases. This innovative database manageability allows database
administrators to spend less time managing routine tasks, and focus on tasks
that help enterprises gain and maintain a sustainable competitive advantage.
For more information on autonomics visit:
http://www.ibm.com/software/data/db2/autonomics/
򐂰 Multi-vendor SQL and application development support
Businesses cannot afford to have their application development teams
struggling with unfamiliar SQL syntax and foreign database APIs. With DB2,
non-DB2 application developers will be happy to notice the full cross-vendor
support for many APIs such as JDBC, ODBC and .NET, as well as rich
support of some other database vendor's SQL, functions, and data types.
http://www.ibm.com/software/data/db2/ad/
򐂰 Web services applications
DB2 can be accessed as a Web service provider, and is usually teamed with
the IBM WebSphere® family products to provide a complete Web services
framework. For more information visit:
http://www.ibm.com/software/websphere/
򐂰 Data warehousing functionality
IBM offers the InfoSphere Warehouse Enterprise Edition, which provides a
powerful range of capabilities that goes beyond traditional warehouses.
InfoSphere Warehouse is a comprehensive platform including tooling and
infrastructure to help data warehouse architects and administrators efficiently
design, deploy and maintain an enterprise data warehouse. For more
information, visit:
http://www.ibm.com/software/data/infosphere/warehouse/

3.1.2 IBM conversion support

With assistance from IBM DB2 experts you can greatly reduce cost, time and
errors associated with the conversion process. DB2 experts are available to

Chapter 3. Planning the conversion from MySQL to DB2 59

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

assist your organization regarding any phase of the conversion process. These
phases include:
򐂰 Assessment of the database and application conversion efforts
򐂰 Project planning
򐂰 System planning
򐂰 Database design
򐂰 Porting preparation and DB2 installation
򐂰 Database structure and data conversion
򐂰 Application conversion
򐂰 Basic Administration
򐂰 Testing and tuning of DB2 data server

Partner conversion projects

IBM PartnerWorld® is designed to help you succeed in the marketplace and
strengthen the relationship with IBM. When enrolled you can utilize
PartnerWorld tools and resource to grow and profit in the market. You can work
with IBM experts to provide you with leading edge support. Visiting the
PartnerWorld you can find information on sales and marketing, technical
resources, and products and technologies. You can also find out about upcoming
events, education opportunities, certifications, products, and a variety of
promotions.

There is no cost to join partner world, you can find more information and register
by visiting the following link:
http://www.ibm.com/partnerworld

If you are a partner and have a conversion project in mind, contact us at

askdata@ca.ibm.com and title your E-mail “MySQL to DB2 migration”.

Customer conversion projects

In more than 3500 cases IBM conversion specialists around the world have
advised customers on conversion projects to DB2. Before starting the
assessment phase, you should contact the Software Migration Project Office
(SMPO) for no charge conversion estimates as well as access to a team of
conversion experts.

If you are a customer and have a conversion project in mind, contact one of the
following contacts according to your geography:
򐂰 In North America and Latin America, contact: db2mig@us.ibm.com
򐂰 In UK, Europe, Middle East and Africa, contact: emeadbct@uk.ibm.com
򐂰 In Japan, India and Asia Pacific, contact: dungi@hkl.ibm.com

60 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

More information about the DB2 conversion team can be found at the Software
Migration Project Office (SMPO) Web site:
http://www.ibm.com/software/solutions/softwaremigration/dbmigteam.html

You can find the most up to date details about current offerings, success stories,
literature, and other information on the DB2 Migrate Now! Web site:
http://www.ibm.com/software/data/db2/migration/

3.1.3 Education
DB2 provides an easy-to-use, feature-rich environment. Therefore, it is important
that those individuals involved in the conversion process be appropriately trained
to take full advantage of its offerings.

There is a lot of training material available such as courses, self-study guides,

and IBM Redbooks publications. IBM also offers a variety of DB2 courses; one
very useful course is the DB2 9.7 Bootcamp. For course details and scheduled
courses near you visit:
http://www.ibm.com/developerworks/wikis/display/im/DB2+9.7+Bootcamp

For further information regarding DB2 training, visit the DB2 Web site at:
http://www.ibm.com/software/data/education/

This DB2 for Linux, UNIX, and Windows conversion Web site can help you find
the information you need to port an application and its data from other database
management systems to DB2. The porting and conversion steps, which are
described in this chapter, appear in the order that they are commonly performed.
In addition to the technical information on this site, IBM customers and IBM
Business partners should check out the Information for IBM customers and
Information for IBM partners links:
http://www.ibm.com/developerworks/db2/zones/porting/partners.html

and
http://www.ibm.com/developerworks/db2/zones/porting/customers.html.

Here you will find additional links and information regarding assistance or
available resources for your port.
http://www.ibm.com/developerworks/db2/zones/porting/index.html

Chapter 3. Planning the conversion from MySQL to DB2 61

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

3.2 Application assessment

An application assessment is the first step in the conversion planning. The
assessment helps to understand the scope of the conversion project and to
prepare a detailed conversion project plan to design the target system.

You need to understand how your application works and what resources are
needed. There are probably a lot of characteristics within your application that
will influence system planning and the scope of the conversion effort.

The main information required for the application assessment includes:

򐂰 Application architecture
– Standalone
– Client/server
– Tier architecture
򐂰 Database architecture
– Number of databases
– Number of tables
– Number of indexes
– Users, access rights and privileges
򐂰 Size of data stored in the database
– Bytes stored in tables
– Bytes stored in indexes
– Log file size
򐂰 Source code language
– PHP
– Perl
– Java
– C/C++
– any other programming language
򐂰 Database interface
– Direct database access through API
– Database access layer such as ODBC/JDBC
򐂰 Operating system
– Linux
– AIX, UNIX
– Windows
– any other operating system
򐂰 Hardware used

62 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

– CPU
– Memory
– Hard disk

In a client/server environment be sure to describe the application in both the

client and server environment.

3.3 System planning

Based on the application profile created during the application assessment, you
should be able to plan your target system properly.

With DB2 support for different hardware platforms and multiple operating
systems such as Linux, Windows, and AIX and so on, platform limitation should
not be an issue. You can select on which system the converted application
should run based on the application nature and future enhancement
requirements.

As shown in Figure 3-1 on page 63, the target system can be the same system
as the source system, or a different one with a different operating system and
hardware. You might even want to make your database server a separate
machine from the machine your application runs on (creating a two-tier
architecture)

Figure 3-1 Sample conversion scenarios

Chapter 3. Planning the conversion from MySQL to DB2 63

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

If you decide to use a new machine for the conversion program, you need to plan
what kind of hardware you want to use, and which operating system you want to
install on it.

In either case you should check if the hardware of your target system meets the
minimum requirements, paying particular attention to the following:
򐂰 Operating system
򐂰 DB2
򐂰 Application
򐂰 Data
򐂰 Conversion tools (if used)

We discuss system requirement in more details in 5.1.1, “System requirements”

on page 88.

3.3.1 Software
You must determine which software must be installed on your target system.
This can include the following:
򐂰 Operating system (Linux, AIX, UNIX, Windows, others)
򐂰 DB2 version
򐂰 Application to be converted
򐂰 Conversion tools (if used and installed on target system)
򐂰 Any software that you have on your source system, which is required by your
application to run properly. This can include, but is not limited to:
– HTTP-server
– Web application server
– Development environment
– Additional software (like LDAP or others)

Be sure to have the latest versions and fix packs of the planned products. Ensure
the chosen software is supported on the chosen operating system.

3.3.2 Hardware
When starting the conversion process it is important to have a target platform
that meets the minimum requirements of all the software that will be installed on
it. Check the supported hardware platforms depending on the chosen software.

Your application also requires hardware resources. Be sure to have enough disk
space for your application and transformed data.

64 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

A great option for operating enablement environments is virtual images. Virtual

images reduce the hardware impact and allow simple sand boxes for the
enablement effort without incurring delta costs. Keep in mind the disk space is
still needed.

3.3.3 Conversion tools

There are free and commercial tools available to assist you in converting your
application from MySQL to DB2. The tools offer a variety of functions and are
supported on various operating systems. IBM offers the free IBM Data
Movement Tool. In addition to this conversion tool IBM also offers an assortment
of free IBM tools, such as the new IBM Optim Data Studio, that can ease the
learning curve by making database administration and debugging queries and
functions effortless.

When deciding to use a tool, be sure that it fulfills the requirements appropriate
for your platform.

IBM Data Movement Tool

The IBM Data Movement Tool offers schema and data movement support from
MySQL, Oracle, Microsoft SQL Server, Microsoft Access, Sybase Adaptive
Server Enterprise, PostgreSQL and Ingres to DB2 9.7. The IBM Data Movement
Tool replaces the Migration Tool Kit with a greatly simplified workflow and high
speed data movement.

The tool can be used to extract the data from the source database into flat files,
generates scripts to create the database objects, and import the data using the
DB2 LOAD utility. At the time this book was written the Data Movement Tool
supports the following database objects for a MySQL to DB2 conversion:
򐂰 Tables, constraints, Indexes, primary keys and foreign keys(with InnoDB)

At the time this book was written the Data Movement Tool does not support the
following database objects for a MySQL database conversion:
򐂰 Views, procedures, functions, triggers, packages

The IBM Data Movement tool is available in both GUI and command line form.
򐂰 Graphical user interface (GUI)
The GUI interface offers the IBM Data Movement Tool conversion
functionality by using a Java interface. It provides an easy to use interface for
beginners.
򐂰 Command line

Chapter 3. Planning the conversion from MySQL to DB2 65

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

The command line interface offers a way to operate the IBM Data Movement
Tool from the command line. The command line interface is intended for
experienced users who want to run end-to-end conversions without user
interaction

For our conversion scenario we use the new IBM Data Movement Tool to convert
database objects and data from MySQL to DB2. If you want to download the IBM
Data Movement Tool or receive more information about it, refer to:
http://www.ibm.com/developerworks/data/library/techarticle/dm-0906datamovement/
index.html

The installation of the IBM Data Movement Tool is covered in Chapter 5,

“Installation” on page 87.

Other available conversion tools

Various other tools are available online, some example are listed below. Note
that IBM does not service these and does not guarantee anything regarding the
usage of these tools:
򐂰 SQLWays
A commercial tool by Ispirer Systems for the Windows, UNIX and Linux
operating system. For more details visit:
http://www.ispirer.com/

3.4 The conversion process

For accurate planning it is important to understand the steps to complete the
conversion project. To estimate the project effort correctly and convert the
database and application successfully, each of the steps should be planned.
Figure 3-2 shows the basic porting process steps.

Figure 3-2 Steps of the conversion process

66 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

3.4.1 Porting preparation and installation

After deciding which target system you want to use, you can set up the following:
installation of hardware components, operating system with users and access
rights, network connections, required software, and finally the DB2 version of
your choice.

For more information on installation of DB2 and the IBM Data Movement Tool,
refer to Chapter 5, “Installation” on page 87.

3.4.2 Database structure porting

Once DB2 is installed in the target, you can port the database structure from your
source MySQL database to DB2.

The database structure is usually described through Data Definition Language

(DDL) statements. DDL scripts include the creation of tables, keys, and indexes.
Since the DDL syntax for MySQL and DB2 is different, a converting process is
required. Using a tool to convert the database structure can save time and effort.
Figure 3-3 shows the conversion database structure basic steps.

Figure 3-3 Database structure porting process

Export DDL from MySQL databases

Information for the tables, keys, indexes, and functions within your source
MySQL database must be retrieved.

Tools such as IBM Data Movement Tool can do most of this task automatically.
Regardless of the tool used, you should verify the output results. This step can
also be performed manually, however, you need to make sure to cover all
database objects.

Chapter 3. Planning the conversion from MySQL to DB2 67

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

Convert DDL to DB2 syntax

Since the DDL used for MySQL is slightly different than the DDL used for DB2,
the DDL used to create MySQL databases and objects must be converted to
DB2 syntax. These syntax differences between MySQL and DB2 are described
in more detail in Chapter 6, “Database conversion” on page 115.

This conversion of DDL syntax is supported by tools such as IBM Data

Movement Tool; or, it may be performed manually as well. In this case watch out
for different data types that are used by MySQL and DB2.

If you plan to change the logical model of your database structure to enhance
your application and take advantage of DB2 functions and features, the DDL
should be modified in this step.

Create DB2 database structure

Once all DDL scripts are converted to DB2 syntax, you can create your DB2
database and execute the DDL scripts to create the database structure.

These actions are supported by tools such as IBM Data Movement Tool. Make
sure to check that all database objects like tables, keys, indexes and functions
are created successfully.

3.4.3 Data porting

With the database installed on your new system and the structure of your
database created, the data extracted should be loaded from the source to the
target system.

Databases usually provide mechanisms for exporting (dumping) and importing

(loading) data. We categorized the MySQL data into:
򐂰 User data: Data that contains information about users, access rights, and
privileges
򐂰 Application data: Data created and used by an application

Figure 3-4 on page 69 shows the basic steps of data movement.

68 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

Figure 3-4 Data porting process

Export MySQL user data

MySQL and DB2 use different database security mechanisms. The database
object access privileges are stored in the MySQL database. You need to
understand what MySQL access rights your application has and how you can
map them to DB2. Depending on your application you probably have to export
user data from the MySQL database.

Map MySQL user data to DB2 user data

Since DB2 users, access rights and privileges are maintained in a different way,
you have to translate the MySQL user data into DB2 commands that grant
privileges and use operation system's user ID management functions to create
the users.

In this step, you create scripts to create the required users and grant them
access privileges to the DB2 database and objects based on the MySQL user
data.

Create DB2 user data

Once you have mapped your MySQL user data to DB2 user data, your users
must be created in the DB2 system and the necessary privileges must be
granted to them.

Chapter 3. Planning the conversion from MySQL to DB2 69

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

The scripts with the DB2 commands for creating users and granting privileges
should be run.

Export MySQL application data

You have to dump the application data out of the MySQL database. This step is
supported by the IBM Data Movement Tool or you can do it manually. Be aware
of the differences in the format of DATE, TIME and other data types.

Convert MySQL application data to DB2 format

You must convert the MySQL application data to a format that can be stored
within DB2. The IBM Data Movement Tool handles the conversion of all MySQL
data types for you.

Import application data into DB2

The exported (and maybe converted) data must finally be loaded into the DB2
tables. The IBM Data Movement Tool also supports this step, and of course data
can be loaded into DB2 manually.

Basic data checking

Once you have loaded all data into your database, basic data checking should be
performed. For example, things like correct number of rows per table and the
correct representation of the field-values must be verified. Also, be sure that you
have all the users created in your DB2 system.

All of the steps described in this section are performed on the MySQL sample
database and explained in great detail in Chapter 7, “Data conversion” on
page 167.

3.4.4 Application porting

Various tools can greatly help the conversion process of a database structure
and its data. The conversion of the application may include some manual
conversions for cases where non-standard SQL functions are utilized. DB2 9.7
introduced a number of features that simplify this task from other relational
database vendors to DB2. Tools exist that support tasks such as syntax
highlighting, but in most cases the main conversion must be done manually. A
good way is simply is performing a search for typical SELECT, INSERT, UPDATE,
and DELETE key words and mark these with specific comments. If the
application supports SQL92 compliant database systems the work is significantly
reduced.

The extent to which you have to change the application code depends on the
database interface that is used in the source application. When a database

70 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

access layer is used, the adoption is not that complicated, otherwise, the effort to
enable the application will likely be higher.

Application source code changes

As there are some differences in the Data Manipulation Language (DML) of
MySQL and DB2, you may have to change SQL statements in your application
code directly.

You might find things that are not natively supported in DB2 that used in MySQL,
so a concept must be established to allow the application to behave in the same
way as prior to conversion.

Database interface
Regardless of the interface used between an application and a database
application, the access to the database must be changed since the database
server has been modified.

If standardized interfaces such as ODBC or JDBC are used, the changes will be
less significant than if the application uses the native API of the database
product.

Condition handling
Depending on the implementation of your application, there might be some
changes in the condition handling part of the application.

Additional considerations
DB2 offers rich, robust functions, which you can take advantage of in your
applications. Some of these features that you may want to consider using in your
application, which are different in MySQL are:
򐂰 Concurrency
򐂰 Locking
򐂰 Isolation level transactions
򐂰 Logging
򐂰 National language support
򐂰 XML support

The steps list in this section are performed with various sample application code
and explained in great detail in Chapter 8, “Application conversion” on page 207.

Chapter 3. Planning the conversion from MySQL to DB2 71

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

3.4.5 Basic administration

Regardless of the conversion, regular maintenance work performed by the
database administrator must still be performed. Any administration issues must
also be taken into account during the conversion planning process.

Every database has its own way for backup and recovery, since these are
common and vital tasks in database administration. The database should be
backed up regularly, and the data retention period should be defined based on
the business requirements.

If you have backup and recovery tasks defined on the source system, you
probably want to convert these tasks as well. Be sure to port any existing scripts
for backup tasks to support DB2.

Both the database backup and recovery functions should be tested to ensure a
safe environment to your application.

Log files
DB2 logs differently than MySQL, so database administrators should be aware of
the logging level that can be set, where log information is stored, and how to read
these logs.

DB2 database administration will be described in detail in Chapter 9, “Database

administration” on page 281.

3.4.6 Testing and tuning

Once your new system is up and running, you have to verify that data and
application functionality have been ported completely, and that system behavior
has not changed in a negative way.

If you succeed with testing you can then proceed to tuning your database and
application in order to speed up your application.

Data checking
Aside from basic data checks that should be performed when exporting and
importing data, checking that your application handles your data correctly and
manipulates the expected fields on inserts, updates, or deletes is a vital part.

This can performance checking can be done manually, or a script may be used
to have the data checked.

72 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch03.fm

Code and application testing

It is very important that the behavior of your application has not changed.
Interactions between components as well as each module of your application
must be tested. A code review of all changed code is recommended as well.

Troubleshooting
Whenever the conversion leads to a problem, such as has wrong data or
incorrect application behavior, you have to determine what the problem is in
order to fix it.

You should understand error messages from the application as well as DB2 error
messages. The troubleshooting process includes studying the DB2 log files.

See the DB2 technical support Web site for help with specific problems:
http://www.ibm.com/software/data/support/

Basic tuning
Once your new system is working perfectly, you might want to tune it for even
better performance. With the correct database configurations, and hints from
DB2 tuning tools, you can speed up your queries quite easily.

DB2 provides tools such as Design Adviser, Performance Monitor or Index

Advisor to support you in speeding up your DB2 system. DB2 also provides the
Self Tuning Memory Manager which keeps your database at an optimal state at
all times without any DBA intervention.

Testing and initial tuning of a DB2 database are discussed in detail in

Chapter 10, “Testing and tuning” on page 321.

Chapter 3. Planning the conversion from MySQL to DB2 73

7093ch03.fm Draft Document for Review October 16, 2009 10:22 am

74 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch04.fm

Chapter 4. Conversion scenario

To illustrate the process of converting a database from MySQL to DB2 we have
developed a small example Web application written in PHP. The backend
database is created using MySQL 5.1 MyISAM tables. This application scenario
is used for a hardware inventory and services tracking application. Although, the
application is written in PHP, it would be possible to convert other popular
programming languages such as Perl, Java and C/C++. Chapter 8, “Application
conversion” on page 207 also provides samples of these popular programming
languages.

The following topics are discussed in detail:

򐂰 Application description
򐂰 Database structure
򐂰 System environment

© Copyright IBM Corp. 2009. All rights reserved. 75

7093ch04.fm Draft Document for Review October 16, 2009 10:22 am

4.1 Application structure

The application used for this scenario is a Web-based inventory management
tracking system. With this application the user can organize and monitor
employee inventory, service tickets, users, and groups.

4.1.1 Application Flow

The flow diagram (Figure 4-1 on page 76) describes the application flow and
functions at a high level.

Figure 4-1 Flow diagram for sample application

76 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch04.fm

When a user enters the Web site, the login page (Figure 4-2) provides three
functions:
򐂰 User Login: To log in as an already registered user
򐂰 New Users: To create a new user account
򐂰 Reset Password: To request the password to a default password

Figure 4-2 Log in page of the sample Web application

With User Login, a registered user uses their user ID and password to log in. The
application verifies the username, password and user permissions against the
registered users in the database. The lowest level of permissions allows the user
to view, edit, and create their inventory and service requests. If the user has
permissions to view, create, and edit groups and users they will be given extra
functionality to do so.

From New Users, a new user can create an account. Completing the registration
form (Figure 4-3 on page 78) creates a new user account in the application. The
application verifies the username provided by the user is unique. By default new
users have the lowest level of permissions. A user who is allowed to edit groups
can add a user to a group with more than default level permissions.

Chapter 4. Conversion scenario 77

7093ch04.fm Draft Document for Review October 16, 2009 10:22 am

Figure 4-3 Registration form of sample application

Using Forget Password, users can reset their account password by entering their
first, last, and user name, as shown in Figure 4-4 on page 79.The new password
will be displayed to the user.

78 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch04.fm

Figure 4-4 Reset password of sample application

Once successfully logged in, the management options are presented to the user.
A set of management options are displayed, and may vary, depending on the
type of permissions held by the logged in user. Figure 4-5 shows the welcome
page options for a user with the highest level of permissions.

Figure 4-5 Welcome/Main menu for administration login of sample application

Using the View/Edit Account Info, users can view account details as shown in
Figure 4-6. Users can update their details by editing the fields and submitting the
form.

Chapter 4. Conversion scenario 79

7093ch04.fm Draft Document for Review October 16, 2009 10:22 am

Figure 4-6 View/Edit account information of sample application

With Add Inventory, the user can associate new inventory with their user
account. Figure 4-7 on page 80 shows a typical filled out Add Inventory form.

Figure 4-7 Add Inventory of sample application

Using View/Edit Inventory List, users have the ability to view their assigned
inventory and other users’ inventory using owner, location, inventory type, or
service created against the inventory (Figure 4-8 on page 81). From this page

80 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch04.fm

users can select to update inventory records by selecting Edit. To see the Edit
button the user must have the permission to update the inventory record.

Figure 4-8 View/Edit inventory list of sample application

Using Create Service Tickets (Figure 4-9 on page 81), users can open service
tickets against their assigned inventory.

Figure 4-9 Create service ticket of sample application

Chapter 4. Conversion scenario 81

7093ch04.fm Draft Document for Review October 16, 2009 10:22 am

From View/Edit Service Tickets, users have the ability to view their created and
assigned service tickets, as shown in Figure 4-10. A user can also view all
service tickets based on user, inventory location, inventory type, and service
type. From this page the user can select to update the service ticket by selecting
Edit. To see the Edit button the user must have the permission to update the
inventory record.

Figure 4-10 View/Edit service tickets of sample application

With View Group Users, users with administration permissions can view all users
within a specific group. From this page the user can update any given user
account by selecting Edit. Figure 4-11 on page 83 shows a user viewing the
manager group user details.

82 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch04.fm

Figure 4-11 View group users of sample application

Using Create/Edit Group (Figure 4-12 on page 83), a user with administration
permissions can create and edit groups.

Figure 4-12 Create/Edit Group of sample application

Chapter 4. Conversion scenario 83

7093ch04.fm Draft Document for Review October 16, 2009 10:22 am

4.2 Database structure

The database that is used for this scenario consists of seven tables, four views,
one trigger, and one stored procedure as described below:
򐂰 The LOCATIONS table contains the building layout information.
򐂰 The OWNERS table contains the user account information for all of the users
that have access to the application. Each owner is associated with a location
from the LOCATIONS table.
򐂰 The INVENTORY table contains the hardware inventory information. Each
piece of inventory is associated with an owner from the OWNER table and a
location from the LOCATIONS table.
򐂰 The SERVICES table contains all service tickets have been open against a
specific piece of hardware inventory from the INVENTORY table. Each
service has one service owner from the SERVICE table associated with the
service. Each service is also associated with a severity level from the
SEVERITY table and a status from the STATUS table.
򐂰 The STATUS table contains the details of the different stages a service ticket
can be in through out its life cycle.
򐂰 The GROUPS table contains group name and permissions.
򐂰 The SEVERITY table contains the degree of importance of getting the service
ticket resolved and the estimated time for each degree.
򐂰 The BOSSGROUP, EMPGROUP, GENERALGROUP, MANGERGROUP,
and TECHGROUP views contain a summary of the user details and user
inventory in each of the groups.
򐂰 The UPDATEDATE trigger can be used to update the closing date field in the
SERVICES table to the current date when the ticket is marked as closed.
򐂰 The UPDATE stored procedure can be used to update the average number of
days it takes to resolve a services ticket in the SEVERITY table. The stored
procedure accepts a specific severity level as an input.

For detailed table information see Figure 4-13 on page 85. We discuss the data
type conversion between MySQL and DB2 in detail in Chapter 6, “Database
conversion” on page 115.

84 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch04.fm

Figure 4-13 Inventory database structure diagram for sample application

4.3 System environment

In this section, we discuss the hardware and software environment used for this
conversion scenario.

This database and application were developed in a virtual machine using

VMware® workstation with the following device setting allocated to the image:
򐂰 Memory: 1020 MB
򐂰 Hard Disk: 40 GB
򐂰 Processors: 2

In our conversion scenario we setup two servers. Our original server which had
the following software installed on the VMware Image:
򐂰 SUSE® Linux 10 SP2
򐂰 MySQL 5.1.36 Community (MySQL AB)
򐂰 Apache 2.0

Chapter 4. Conversion scenario 85

7093ch04.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 PHP 5.3.0

The second VMware image, the destination server, has the following software
installed on the VMware image:
򐂰 SUSE Linux 10 SP2
򐂰 DB2 9.7 Express-C
򐂰 Apache 2.0
򐂰 PHP 5.3.0
򐂰 IBM Data Movement Tool

For more information about VMware workstation and working with VMware
images go to:
http://vmware.com/

DB2 9.7 Express-C for Linux, UNIX and Windows can be downloaded from:
http://www.ibm.com/software/data/db2/express/ .

IBM Data Movement Tool is used to simplify and greatly decrease time it takes
to convert from MySQL to DB2. This tool is available free of charge from IBM at
the following URL:
http://www.ibm.com/developerworks/data/library/techarticle/dm-0906datamovement/
index.html

With IBM Data Movement Tool, the conversion of database objects such as
tables and data types, and the conversion of data can be done automatically into
equivalent DB2 database objects.

The installation and configuration of DB2 and the IBM Data Movement Tool is
covered in the next chapter.

86 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

Chapter 5. Installation
In this chapter we discuss the target system environment setup. For the
database server, we guide you through the installation process of DB2 9.7 for
Linux including the hardware and software prerequisites. The application server
has to be examined to ensure that the exiting software has proper DB2 support.
If this is a completely new system setup, make sure that all the required software
is included in the installation list. Furthermore, we describe the download and
steps required to setup the IBM Data Movement Tool (DMT).

© Copyright IBM Corp. 2009. All rights reserved. 87

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

5.1 DB2 Express-C 9.7 on Linux

Before you start your conversion project it is important to ensure that the system
you choose meets the necessary operating system, hardware, software and
communication requirements.

5.1.1 System requirements

This section provides information about supported Linux distributions, hardware,
software, and communication requirements for DB2 on Linux.

Linux distributions supported by DB2

Table 5-1 lists some IBM DB2 validated and recommended Linux distributions at
the time of writing this book. For the most recent, all-supported Linux
distributions that have gone successfully through the IBM DB2 for Linux
validation program, check:
http://www.ibm.com/software/data/db2/linux/validate/

Table 5-1 Currently supported Linux distributions, kernels, and libraries

Platform Distribution Kernel Library Comment

x86 Red Hat® 2.6.18- libstdc++.so.

Enterprise 92 5
Linux (RHEL) 5

SUSE Linux 2.6.27. glibc-2.9-13.2

Enterprise 19-5
Server (SLES)
11

SUSE Linux 2.6.16 glibc-2.4-31

Enterprise
Server (SLES)
10 SP2

Ubuntu 8.0.4 2.6.24- glibc-2.7.so

19

x86_64 Red Hat 52.6.18 libstdc++.so -all 32-bit compatibility

AMD64/ Enterprise -92l libraries are required for
EM64T Linux (RHEL) 32-bit applications.

88 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

SUSE Linux 2.6.27. glibc-2.9-13.2 -all 32-bit compatibility

Enterprise 19-5 libraries are required for
Server (SLES) 32-bit applications.
11

SUSE Linux 2.6.16 glibc-2.4-31 -all 32-bit compatibility

Enterprise libraries are required for
Server (SLES) 32-bit applications.
10 SP2

Ubuntu 8.0.4 2.6.24- glibc-2.7.so

19

Hardware requirements
DB2 products are supported on the following hardware:
򐂰 HP-UX
– Itanium® based HP Integrity Series Systems
򐂰 Linux
– x86 (Intel Pentium®, Intel Xeon®, and AMD) 32-bit Intel and AMD
processors
– x64 (64-bit AMD64 and Intel EM64T processors)
– POWER® (IBM eServer™ OpenPower®, iSeries, pSeries®, System i,
System p, and POWER Systems that support Linux)
– eServer System z or System z9®
򐂰 Solaris
– UltraSPARC or SPARC64 processors
– Solaris x64 (Intel 64 or AMD64)
򐂰 Windows
– All Intel and AMD processors capable of running the supported Windows
operating systems (32-bit and 64-bit base systems)

For more information on DB2 9.7 system requirements and other DB2 release
system requirement, check:
http://www.ibm.com/software/data/db2/9/sysreqs.html

Disk requirements for DB2 servers

The disk space required for your product depends on the type of installation you
choose and the type of file system you have. The DB2 Setup wizard provides
dynamic size estimates based on the components selected during a typical,

Chapter 5. Installation 89
7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

compact, or custom installation. The following are the sizes for DB2 Express-C
installation:
򐂰 Typical: Requires 500 to 610 MB
With Typical installation type, DB2 is installed with most of the features and
functionality, including graphical tools such as the Control Center and DB2
Instance Setup wizard.
򐂰 Compact: Requires 450 to 550 MB
With Compact installation type, only basic DB2 features and functions are
installed, minimal configuration will be performed and graphical tools are not
included.
򐂰 Custom: Requires 450 to 1080 MB
With Custom installation type, you can select the features you want to install.
The disk space needed varies based on the selected features.

When you install DB2 Enterprise Server Edition or Workgroup Server Edition
using the DB2 setup wizard, size estimates are dynamically provided by the
installation program based on installation type and component selection.

If the space required for the installation type and components exceeds the space
found in the path specified, the setup program issues a warning about insufficient
space. The installation is allowed to continue. If the space for the files being
installed is in fact insufficient, installation will stop, and the setup program will
need to be aborted if additional space cannot be provided.

Remember to include disk space for required databases, software and

communication products.

On Linux and UNIX operating systems, 2 GB of free space in the /tmp directory is
recommended.

Memory requirements for servers

At minimum, a DB2 database system requires 256 MB of RAM. For a system
running a DB2 product and the DB2 GUI tools, a minimum of 512 MB of RAM is
required. For improved performance, 1 GB of RAM is recommended. These
requirements do not include additional memory requirements for other software
components running on your system.

When determining memory requirements, ensure to remember the following:

򐂰 DB2 products that run on HP-UX Version 11i for Itanium-based™ systems
require 512 MB of RAM at a minimum.

90 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

򐂰 For IBM data server client support, these memory requirements are for a
base of five concurrent client connections. You will need an additional 16 MB
of RAM per five client connections.
򐂰 Memory requirements are affected by the size and complexity of your
database system, as well as by the extent of database activity and the
number of clients accessing your system.
򐂰 For DB2 server products, the self-tuning memory feature simplifies the task of
memory configuration by automatically setting values for several memory
configuration parameters. When enabled, the memory tuner dynamically
distributes available memory resources among several memory consumers
including sort memory, the package cache, lock list memory, and buffer
pools.
򐂰 Additional memory may be required for non-DB2 software that may be
running on your system.
򐂰 Specific performance requirements may determine the amount of memory
needed.
򐂰 On Linux operating system, SWAP space at least twice as large as RAM is
recommended.

Communication requirements
When using TCP/IP as the communication protocol, no additional software is
needed for connectivity. For more supported communication protocols, refer to
DB2 manual Quick Beginnings for DB2 Servers 9.5, GC10-4246 at:
http://publibfp.boulder.ibm.com/epubs/pdf/c2358642.pdf

or visit the IBM DB2 Database for Linux, UNIX and Windows Information Center
at:
https://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

5.1.2 Installation procedure

Table 5-2 shows the four installation methods in which DB2 can be installed. For
completeness we have provided the information for supported Linux, UNIX, and
Windows operating systems. Each of the methods listed has its own pros and
cons, and varies upon the environment. For a discussion on the preferred
method, refer to the Up and Running with DB2 on Linux, SG24-6899.

Table 5-2 DB2 installation methods

Installation method Linux UNIX Windows

DB2 setup wizard yes yes yes

Chapter 5. Installation 91
7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Installation method Linux UNIX Windows

db2_install yes yes no

Response file installation yes yes yes

Payload file deployment yes yes no

(Manual installation)

For this particular project or any other conversion project in general, the DB2
Data Server Client is required. It provides libraries for application development. If
the application server and the database server are to be placed at same system,
you can install both the DB2 server and Data Server Client in one step by
selecting Custom installation type.

For this project we perform the following steps to install DB2 9.7 on Linux:
1. Log on to Linux as a root user.
2. Download DB2 9.7 Express-C from:
http://www.ibm.com/software/data/db2/express/download.html
3. Save the tar file to /usr/local/src directory.
4. Change directories to the /usr/local/src directory:
cd /usr/local/src/expc
5. Extract the tar file:
tar -xzf db2exc_970_LNX_x86.tar.gz
6. Change to the directory:
cd /usr/local/src/expc
7. Launch DB2 Setup wizard with which opens Figure 5-1 on page 93:
./db2setup

92 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

Figure 5-1 DB2 setup launch pad

Note: Starting in DB2 9.5 you can also do non-root installation of DB2.

8. Choose Install a Product once the DB2 Launchpad opens as shown in

Figure 5-2 on page 94.

Chapter 5. Installation 93
7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Figure 5-2 DB2 setup launch pad

9. Select Install New under the option to install the server to launch the DB2
setup wizard.
10.Go to the Software License Agreement panel and read the Software License
Agreement, as shown in Figure 5-3 on page 95. If you agree with the
agreement select the Accept and click Next to go to the next screen.

94 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

Figure 5-3 DB2 software license agreement panel

11.In the Installation Type panel click Custom as shown in Figure 5-4 on page 96
and click Next.

Chapter 5. Installation 95
7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Figure 5-4 DB2 installation type panel

12.Click Next again to get the Features panel. Select Application Development
tools option, as shown in Figure 5-5 on page 97, and click Next.

96 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

Figure 5-5 DB2 custom installation with application development tools selected

13.In the Languages panel choose the type of languages to install and click
Next.
14.In the Documentation panel choose where to access the DB2 Information
Center. You can choose to install it as part of this process or you can access
the online DB2 Information Center at any time. Click Next.
15.In the Database Administration Server (DAS) panel fill out the DAS user
information. Linux group and user accounts do not have to be created prior to
this step; DB2 will create the required Linux system group and user
automatically. For the example installation, we use the default name dasusr1
and choose a password for this user, as shown in Figure 5-6 on page 98 and
click Next.

Chapter 5. Installation 97
7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Figure 5-6 DB2 setup administration server window

16.In the Instance setup panel, you can choose whether you would also like to
setup an Instance during the DB2 installation. By selecting Create a DB2
instance and clicking Next we let DB2 create the instance for us.
17.Fill out the instance owner information in the Instance Owner panel. Linux
group and user accounts do not have to be created prior to this step; DB2 will
create the Linux group and user. For the example installation, we use the
default db2inst1 settings and create a password for this user, as shown in
Figure 5-7 on page 99 and click Next.

98 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

Figure 5-7 DB2 setup DB2 instance owner

18.In the Fenced user panel, we allow DB2 create the ID for us, as shown in
Figure 5-8 on page 100, and click Next.

Chapter 5. Installation 99
7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Figure 5-8 DB2 setup DB2 fenced user

19.In Instance Communication panel, configure the DB2 instance TCP/IP

communication. For the example installation, we use the default settings and
click Next.
20.In the DB2 Tools Catalog Configuration panel, we create a local tools catalog
by selecting the db2inst1 instance. We also use the default local database
TOOLSDB and default schema SYSTOOLS, then click Next. Figure 5-9 on
page 101 shows the DB2 Tools Catalog Configuration panel.

100 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

Figure 5-9 Prepare DB2 tools catalog

21.At the end, the setup wizard provides a summary of the installation options
selected. Review it and click Finish to start the installation.

Fix pack installation

We recommend you install the latest DB2 Fix Pack.
1. Download Fix Pack from:
http://www.ibm.com/software/data/db2/support/db2_9/
2. Change to the directory in which the installation image is located.
3. Enter the installFixPak command to launch the installation.
4. Update instances to run against the new code with db2iupd.
5. Update DB2 Administration Server (DAS) using dasupdt.
6. Restart applications.

db2setup command options

The db2setup command provides options for specifying the locations of the log
files, trace file, or response file created during installation (see Figure 5-10 on
page 102). Log files are useful in verifying the installation status and tracing the

Chapter 5. Installation 101

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

problem if installation failed. By default db2setup starts the graphical user

interface (GUI) and picks up the language flag from the operating systems
settings.

Figure 5-10 db2setup command options

If the log file option is not specified, the db2setup.log and db2setup.err are stored
in the /tmp directory on a Linux operating system. Example 5-1 shows an
example of the db2setup.log.

Example 5-1 DB2setup log file

DB2 Setup log file started at: Fri Jul 24 15:20:15 2009 EDT
============================================================

Operating system information: Linux 2.6.16.60-0.21-smp.#1 SMP Tue May 6

12:41:02 UTC 2008 i686

Product to install: DB2 Express-C

Installation type: Custom
...

5.1.3 Instance creation

For our project we allow DB2 to create the DB2 instance automatically during
installation (Figure 5-8 on page 100). Although this option is the default, you can
turn automatic instance creation off during installation and create instances and
databases manually after the installation has completed.

If you have chosen not to create instances during the DB2 installation or need
add another instance after installation, there are two options to create instances
manually.
򐂰 The db2isetup command starts a graphical tool for creating and configuring
instances as shown in Figure 5-11 on page 103. It allows you to specify all the
required configuration parameters such as the instance owner and
communication protocol in an easy and guided fashion. The command can be
found in /opt/ibm/db2/V9.7/instance on a Linux operating system.

102 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

Figure 5-11 Graphical user interface for db2isetup

򐂰 The second option is the db2icrt command. It is a command-line alternative

to create the instances. For example:
db2icrt -u db2fenc1 db2inst1

The command to create the DAS user is dascrt and is used in the following way:
dascrt -u dasadm1

As part of the GUI instance creation, the installer suggests three users identified
as db2inst1, db2fenc1, and dasadm1. These are default names for the instance
users. If you do not want to use the default names, you can choose your own by
creating the system user IDs and groups ahead of time and inputting these
parameters in the wizard when prompted. The installer will also add the following
entry to the /etc/services file in order to allow communication from DB2 clients:
db2c_db2inst1 50000

Where db2c_db2inst1 indicates the service name, and 50000 indicates the port
number. DB2 allows for multiple instances on one server installation to allow for
various environments, that is, test, production, development, and so on.
Subsequent instances may be created on the same server simply by using one
of the methods introduced above.

5.1.4 Client setup on Linux

This section discusses installation and configuration of DB2 clients to access a
remote DB2 server. DB2 provides two types of clients free of charge:
򐂰 DB2 Runtime Client. This client is best suited for enabling applications to
access DB2 servers.
򐂰 DB2 Client. This client includes all the functionality found in the DB2 Runtime
Client plus functionality for client-server configuration, database
administration, and application development.

Chapter 5. Installation 103

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

All clients are supported on Linux, AIX, HP-UX, Solaris, and Windows operating
systems.

Note: Typically, in a production environment, DB2 clients are installed on

separate physical machines from the DB2 server. However, for an application
development environment, it may be useful to have everything, such as the
DB2 database server plus the clients, on the same machine.

To access a remote DB2 database, you can either run the easy to use graphical
tool Configuration Assistant or use the catalog commands to provide entries for
the following three directories:
򐂰 NODE directory: A list of remote DB2 instances
򐂰 ADMIN NODE directory: A list of remote DB2 Administration servers
򐂰 DATABASE directory: A list of databases

To use the command-line tools, first catalog the DB2 node. The DB2 node is the
server where the database resides, then catalog the database. See
Example 5-2.

Example 5-2 Cataloging node and database

--
-- catalog database node
--
CATALOG TCPIP NODE db2node REMOTE server1 SERVER 50001

--
-- catalog the DAS on the remote node
--
CATALOG ADMIN TCPIP NODE db2das remote SERVER1
-
-- catalog database
--
CATALOG DATABASE invent AS inventdb AT NODE db2node

After installing your DB2 Client, you should configure it to access a remote DB2
server using the Configuration Assistant. The graphical interface can be
launched through the DB2 Control Center or ran on its own by using the
command db2ca. For more details refer to the IBM DB2 manual Quick
Beginnings for DB2 Clients, GC10-4242.

104 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

5.2 Other software product

All software requirements for the target environment should be identified in the
conversion planning and preparation stage. Prior to conversion of any data
objects, all software must be installed and configured.

The sample application in this book is written in PHP with Apache2. Therefore,
the next two sections discuss how to prepare the target system for Apache2 and
PHP.

5.2.1 Apache2 installation with DB2 support

When converting an application from one server to another, you must ensure that
all software is properly installed on the new server. In our conversion scenario
we use Apache.

Installation steps
The following steps explain how we installed Apache2 on SUSE 10 SP2:
1. Downloading the Apache package.
The source code for Apache package is available at:
http://httpd.apache.org/download.cgi
In our conversion scenario we used Version 2.2.11 of Apache, and the
package we downloaded was httpd-2.2.11.tar.gz.
2. Changing the working directory.
Use the cd command to make your working directory the directory you
downloaded the tar file to:
db2server: # cd /usr/local/src/
3. Uncompress the source package.
The following command decompress the contents of the source package into
a directory called httpd-2.2.11:
db2server:/usr/local/src # tar -xzf httpd-2.2.11.tar.gz
4. Changing the working directory.
Use the cd command to make the newly created directory your working
directory:
db2server:/usr/local/src # cd httpd-2.2.11/
5. Specifying the configuration options for the PHP source.

Chapter 5. Installation 105

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Possible configuration options can be listed by issuing the following

command:
db2server:/usr/local/src/httpd-2.2.11# configure -help
6. Run the configure script.
The configure script builds the Makefile. The following is the configuration
command we used to setup our server:
db2server:/usr/local/src/httpd-2.2.11# ./configure
--prefix=/usr/local/apache2
--enable-so
--enable-cgi
--enable-info
--enable-rewrite
--enable-speling
--enable-usertrack
--enable-deflate
--enable-ssl
--enable-mime-magic
Where option:
--prefix specifies the Apache install directory
--enable-so dynamically Shared Object (DSO) capability, which
allows you to load modules into Apache at runtime
--enable-cgi enables support for CGI scripts,
--enable-info enables server information
--enable-rewrite enables rule based URL manipulation
--enable-speling when enabled apache will corrects common URL
misspellings
--enable-usertrack enables user-session tracking
--enable-deflate deflate transfer encoding support
--enable-ssl SSL/TLS support (mod_ssl)
--enable-mime-magic automatically determining MIME type
You can run the configure command with the --help to get a full list of
options for configuring Apache
7. Compiling Apache
After configuring the source files, the compile process is started using the
make command. We piped the output to a log file in order to check for failure
afterwards:
db2server:/usr/local/src/httpd-2.2.11 # make > apacheMake.out

106 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

8. Installing Apache
Once Apache has compiled successfully, it can be installed as the root user:
db2server:/usr/local/src/httpd-2.2.11 # make install
9. Add the apachect1 script to the following directories.
Use the ln command to create a link to the apachect1 file in the /usr/bin and
/etc/init.d directories:
ln -s /usr/local/apache2/bin/apachectl /usr/bin/apachectl
ln -s /usr/local/apache2/bin/apachectl /etc/init.d/.
10.Starting Apache httpd server.
Use the following command to start the apache httpd server:
apachectl start

5.2.2 PHP installation with DB2 support

There are four main interfaces available for connecting to a DB2 database in
PHP:
򐂰 The ibm_db2 extension interface provides an API to read/write from/to the
database. This extension makes it easy to convert Unified ODBC applications
to use ibm_db2.
򐂰 PDO_ODBC and PDO_IBM are object-oriented extension interfaces that are
used to connect to a database. Both are based on PDO (PHP Data Object)
standards, PDO_IBM includes the IBM database driver for PDO.
򐂰 Unified ODBC is traditional procedural interface and supports multiple
database servers. The Unified ODBC extension is not optimized for DB2 and
therefore not recommended for new applications.

When converting a PHP application on an existing server, some preparation and

verification must be performed to have DB2 connection support. If a precompiled
package of PHP is used, by default only a predefined set of functions and
database support is integrated. To see which configuration options are included
in the previous installed PHP version, the PHP function phpinfo() can be invoked
out of a Web page file. In the third section of the PHP information, the included
configuration commands are listed as shown in Figure 5-12 on page 108.

Chapter 5. Installation 107

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Figure 5-12 PHP Configuration options

Note: All commands and procedure descriptions provided in this section refer
to SUSE Linux Enterprise Server 10 SP2. This can vary for other versions or
Linux distributions.

Installation steps:
In order to use the IBM DB2 libraries, PHP has to be recompiled. These
installation steps explain how to update your PHP install and install PHP from
scratch:
1. Backup the httpd.conf and php.ini files.
To ensure the configuration files for Apache and PHP are not lost when
installing the new PHP version, we recommend backing up the
/etc/httpd/httpd.conf and if you have a previous version of PHP installed,
back up the /etc/php.ini files.
2. Downloading the PHP package.
The source code for PHP is available at:
http://www.php.net/downloads.php

108 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

In our conversion scenario we used Version 5.3.0 of PHP, and the package
we downloaded was php-5.3.0.tar.gz.
Download the ibm_db2 PECL extension at:
http://pecl.php.net/package/ibm_db2
In our conversion scenario we used Version 1.2.8, and the package we
downloaded was ibm_db2-1.8.2.tgz.
Download the PDO_IBM PECL extension at:
http://www.pecl.php.net/package/PDO_IBM
In our conversion scenario we used Version 1.3.0, and the package we
downloaded was PDO_IBM-1.3.0.tgz.
3. Uncompress the source package.
The following command decompresses the contents of the sources package
in to a directory called php-5.3.0:
db2server:/usr/local/src # tar xzf php-5.3.0
4. Add the PECL extensions to php install directory.
Use the mv command to move the compressed files to the extension directory
in the install directory:
db2server:/usr/local/src # mv ibm_db2-1.8.2.tgz php-5.3.0/ext/.
db2server:/usr/local/src # mv PDO_IBM-1.3.0.tgz php-5.3.0/ext/.
5. Changing the working directory.
Use the cd command to make the ext directory your working directory:
db2server:/usr/local/src # cd php-5.3.0/ext/
6. Uncompressing PECL extension packages.
The following command decompresses the contents of the extension
packages into directories called ibm_db2-1.8.2 and PDO_IBM-1.3.0:
db2server:/usr/local/src/php-5.3.0/ext/ # gzip -d < ibm_db2-1.8.2.tgz |
tar -xvf -
db2server:/usr/local/src/php-5.3.0/ext # gzip -d < PDO_IBM-1.3.0.tgz |
tar -xvf -
7. Rename the extension directories.
Use the mv command to rename the extension directories:
db2server:/usr/local/src/php-5.3.0/ext # mv ibm_db2-1.8.2 ibm_db2
db2server:/usr/local/src/php-5.3.0/ext # mv PDO_IBM-1.3.0 pdo_ibm
8. Changing the working directory.
Use the cd command to make the PHP install directory your working
directory:

Chapter 5. Installation 109

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

db2server:/usr/local/src # cd /usr/local/src/php-5.3.0/
9. Remove the configure file.
Use the rm command to remove the PHP configure script:
db2server:/usr/local/src/php-5.3.0 # rm configure
10.Rebuild the configure file.
The rebuildconf command rebuilds the configure file to include the new
extensions:
db2server:/usr/local/src/php-5.3.0 # ./buildconf --force
11.Verify the extension is now with in the configure file.
Use the following command to verify the buildconf worked successfully:
db2server:/usr/local/src/php-5.3.0 # ./configure --help | grep
with-ibm-db2
db2server:/usr/local/src/php-5.3.0 # ./configure --help | grep pdo-ibm
12.Specify the configuration options for the PHP source.
A list of the possible configuration options can be seen by issuing the
following command:
db2server:/usr/local/src/php-5.3.0 # configure -help
13.Run the configure script.
The configure script builds the Makefile. For our purposes we specified the
configure command as follows:
db2server:/usr/local/src/php-5.3.0 # ./configure
--prefix=/usr/local/apache2/php
--with-IBM_DB2=/opt/ibm/db2/V9.7
--with-pdo-ibm=/opt/ibm/db2/V9.7
--with-pdo-odbc=ibm-db2,/home/db2inst1/sqllib
--with-ibm-db2=/opt/ibm/db2/V9.7
--with-apxs2=/usr/local/apache2/bin/apxs
--with-config-file-path=/usr/local/apache2/php
Where option:
--prefix specifies the PHP install directory
--with-IBM_DB2 is for the ibm_db2 extension
--with-pdo-ibm is for the PDO_IBM extension
--with-pdo-odbc is for the PDO_ODBC extension
--with-ibm-db2 is for the Unified ODBC extension
--with-apxs2 is for the Apache apxs tool, which allows you to
build extension modules to add to Apache’s
functionality

110 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

--with-config-file-path allows you to specify the path to the php.ini file.

If a particular extension is not required it can be removed from the configure
command and install.
14.Compile PHP.
After configuring the source files, the compile process is started using the
make command. We piped the output to a log file in order to check for failures
afterwards:
db2server:/usr/local/src/php-5.3.0 # make > phpMake.out
15.Install PHP.
Once PHP has compiled successfully, it can be installed as the root user:
db2server:/usr/local/src/php-5.3.0 # make install
16.Configure Apache.
If Apache was already configured for PHP on your server, only some changes
have to be made. As part of the installation process, PHP automatically
modifies Apache configuration file /usr/local/apache2/conf/httpd.conf.
Confirm that the httpd.conf has the following lines. If the lines are missing,
update the file.
– LoadModule php5_module modules/libphp5.so
– directoryIndex index.php index.html index.htm index.html.var
– AddType text/html php
– AddType application/x-httpd-php.php
17.Check php.ini.
Either copy php.ini-dist in the Configuration File Path (see Figure 5-12 on
page 108) of PHP and rename it to php.ini, or if your PHP version is the same
as before, copy the version that was saved in step 1 back in this directory.
18.Start Apache httpd server.
Apache has to be restarted to use the configuration changes:
db2server:/usr/local/src/php-5.3.0 # apachect1 restart

5.3 IBM Data Movement Tool installation and usage

The IBM Data Movement Tool is an ingenuous tool that offers schema and data
movement from various database providers to DB2. The tool accepts information
about the source database and target database, extracts the database objects
and data from source database, and can input this data into the target DB2
database.

Chapter 5. Installation 111

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

For the IBM Data Movement user guide and a free download of the IBM Data
Movement Tool visit:
http://www.ibm.com/developerworks/data/library/techarticle/dm-0906datamovement/

5.3.1 IBM Data Movement Tool prerequisites

The following prerequisites are required to successfully deploy the IBM Data
Movement Tool for MySQL:

Software requirements
The software required to use the IBM Data Movement Tool are described in this
topic.

General requirements
In general, you should have the following:
򐂰 Latest version of the IBM Data Movement Tool.
򐂰 MySQL: Ensuring that MySQL is running (usually the daemon can be started
with the command safe_mysqld & from an account with root permissions).
򐂰 DB2 V9.7 should be installed on the target server. Use the command
db2start to ensure the DB2 Server is up and running.
򐂰 Java version 1.5 or higher must be installed on your target server. To verify
your current Java version, run java -version command. By default, Java is
installed as part of DB2 for Linux, UNIX, and Windows in
<install_dir>\SQLLIB\java\jdk (Windows) or /opt/ibm/db2/V9.7/java/jdk
(Linux).
򐂰 You must have the JDBC drivers for the MySQL source database
(mysql-connector-java-5.1.8-bin.jar or latest driver) and the DB2 target
database (db2jcc.jar, db2jcc_license_cu.jar or db2jcc4.jar,
db2jcc4_license_cu.jar) installed on the server with the IBM Data Movement
Tool.

Operating system
DMT supports the following operating system:
򐂰 Windows
򐂰 z/OS
򐂰 AIX
򐂰 LINUX
򐂰 Solaris
򐂰 HP-UX
򐂰 Mac

112 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch05.fm

For the purpose of this document we have used the IBM Data Movement Tool
with DB2 Version 9.7 and MySQL Version 5.1.36. An installation on the DB2
server side is recommended to achieve the best data movement performance.

5.3.2 IBM Data Movement Tool installation

Installing the IBM Data Movement Tool is simple and straight forward:
1. Download the latest version of the IBM Movement Tool from:
http://www.ibm.com/developerworks/data/library/techarticle/dm-0906datamovem
ent/
2. Create a directory for the IBM Data Movement Tool and copy the
IBMDataMovementTool.zip to this directory.
db2server:/opt/ibm # mkdir IBMDataMovementTool
db2server:/usr/local/src # cp IBMDataMovementTool.zip
/opt/ibm/IBMDataMovementTool/.
3. Expand the compressed package using:
db2server:/opt/ibm/IBMDataMovementTool/ # unzip IBMDataMovementTool.zip
To launch the IBM Data Movement Tool execute command
IBMDataMovemtnTool.sh for Linux and UNIX and IBMDataMovmentTool.cmd for
Windows. Figure 5-13 on page 114 shows the IBM Data Movement Tool
screen.

Chapter 5. Installation 113

7093ch05.fm Draft Document for Review October 16, 2009 10:21 am

Figure 5-13 IBM Data Movement Tool

114 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Chapter 6. Database conversion

After the conversion plan has been created and all required software has been
installed and set up, it is time to translate the source MySQL database structure
into DB2.

In this chapter we discuss the process of converting the database structure from
MySQL 5.1 server to DB2 9.7 server. Before doing this we must look into the
differences between MySQL and DB2 database structure.

In the first section we discuss data type mapping, taking a closer look at MySQL
and DB2 data types and the differences between them. Following this section,
Data Definition Language (DDL) differences are described providing a basic
syntax comparison between MySQL and DB2.

In the succeeding section we provide additional considerations for users to keep

in mind while porting the database schema from MySQL to DB2.

The last section provides detailed information regarding database schema

porting steps using the following approaches:
򐂰 Porting using the IBM Data Movement Tool
򐂰 Manual porting
򐂰 Metadata transform

© Copyright IBM Corp. 2009. All rights reserved. 115

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

6.1 Data type mapping

In this section we compare the differences between MySQL and DB2 data types.
In general, all MySQL data types can be mapped to DB2 data types. With the
assistance of the IBM Data Movement Tool this process can be effortless.

Every column in database table has an associated data type which determines
the values that this column could contain. DB2 supports both built-in data types
and user-defined data types (UDT) whereas MySQL only supports built-in data
types. Figure 6-1 shows built-in data type of MySQL.

Figure 6-1 MySQL data types

Figure 6-2 on page 117 shows built-in data types supported by DB2.

116 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Figure 6-2 DB2 data types

MySQL data types are grouped into three categories, and can be converted to
DB2 data types following the rules suggested as follows:
򐂰 Numeric type
– TINYINT
This is a single byte integer in MySQL that can be mapped to a DB2
SMALLINT for similar functionality.
– SMALLINT
A small integer is a two-byte integer with a precision of five digits. With
MySQL, the range of signed small integers is (-32768 to 32767) making it
replaceable by DB2 SMALLINT. For unsigned MySQL small integers the
range is (0 to 65535) making it replaceable by DB2 INTEGER.
– BIT, BOOL, and BOOLEAN
These are synonyms for TINYINT(1). Instead of BIT, BOOL, and
BOOLEAN, DB2 uses SMALLINT with check constraint.
– MEDIUMINT

Chapter 6. Database conversion 117

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

This is a medium-size integer with a signed range of (-8388608 to

8388607) or (0 to 16777215) for unsigned. DB2 uses an INTEGER for this
which has a range of (-2147483648 to 2147483647).
– INTEGER and INT
An integer is 4-byte integer for both MySQL and DB2. The range of a
signed MySQL INTEGER and DB2 INTEGER is (-2147483648 to
2147483647), therefore, the DB2 INTEGER can be used. For unsigned
MySQL INTEGER the range is 0 to 4294967295. This can be replaced by
DB2 BIGINT.
– BIGINT
A big integer is an 8-byte integer for both MySQL and DB2. A DB2 BIGINT
can be used for a signed MySQL BIGINT because the range of a signed
MySQL BIGINT and a DB2 BIGINT is (-9223372036854775808 to
9223372036854775807). For unsigned MySQL BIGINT the range is (0 to
18446744073709551615), DB2 DECIMAL will cover the range.
– FLOAT
A FLOAT in MySQL is single precision floating-point number ranging from
-3.402823466E+38 to -1.175494351E-38, 0, and 1.175494351E-38 to
3.402823466E+38. Whereas in DB2 it is a double-precision floating-point
number ranging from -1.79769E+308 to -2.225E-307, or from 2.225E-307
to 1.79769E+308. Hence, a FLOAT value in MySQL can be directly
mapped to a DOUBLE in DB2.
– DOUBLE
A DOUBLE is a double precision floating-point number for both DB2 and
MySQL. For signed MySQL DOUBLE is can be directly mapped to
DOUBLE in DB2. Unsigned MySQL DOUBLE can be mapped to
DECIMAL in DB2.
– REAL
This is synonym for DOUBLE in MySQL, therefore can be mapped to the
same DB2 data types as the MySQL DOUBLE data type.
– DECIMAL, NUMERIC, and FIXED
A DECIMAL in MySQL is mapped to a DECIMAL in DB2. Although MySQL
and DB2 implement decimals in a different way, externally they behave
same.
򐂰 Date and time type
– DATE
A DATE in MySQL and DB2 DATE both use four bytes (first two for year,
third for month, and last for day). The range of the MySQL date for year is

118 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

1000-9999, whereas DB2 supports date from 0001-9999, allowing DB2 to

map this MySQL data type.
– DATETIME
A date and time combination in MySQL is displayed as YYYY-MM-DD
HH:MM:SS ranging from year 1000 to 9999. In DB2, TIMESTAMP is used
for a similar purpose, which is a seven part value (year, month, day, hour,
minute, second, and microsecond).
– TIMESTAMP
A timestamp in MySQL is a date/time combination with range of
1970-01-01 00:00:00 to the year 2037. It is automatically set to the date
and time of the most recent operation if you do not give it a valid value.
This data type can be mapped to the DB2 TIMESTAMP or an optional
choice is the DB2 TIME data type.
– TIME
MySQL time represents a clock ranging from -838:59:59 to 838:59:59. It is
mapped to DB2 TIME, which is a 24 hour clock.
– YEAR
MySQL year can be in two or four-digit format representing the year from
1901 to 2155 mapped to SMALLINT or CHAR(4) in DB2.
򐂰 String and character types
– CHAR
A fixed length string in MySQL is represented with the same name in DB2
mapping to a CHAR in DB2.
– VARCHAR
A variable-length string in MySQL has a maximum length of 65,535 and a
DB2 variable-length string has a maximum length of 32,672. A MySQL
VARCHAR can be mapped to a DB2 VARCHAR when the MySQL
VARCHAR variables are less than 32, 672; otherwise, VARCHAR can be
mapped to a CLOB.
– BINARY
MySQL BINARY stores binary byte strings which can be mapped to DB2
CHAR(I).
– VARBINARY
MySQL VARBINARY stores binary byte strings which can be mapped to
DB2 VARCHAR(I).
– TINYBLOB

Chapter 6. Database conversion 119

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

MySQL TINYBLOB is a binary large object column with a maximum length

of 255 which can be mapped to DB2 BLOB(255).
– TINYTEXT
MySQL TINYTEXT is a character stream of maximum length 255I which
can be mapped to DB2 CLOB(255).
– BLOB
MySQL BLOB is a binary data column with a maximum length of 65535
which can be mapped to DB2 BLOB(65K).
– TEXT
This is a TEXT column with a maximum length of 65535 and is mapped to
DB2 CLOB(65K).
– MEDIUMBLOB
MySQL MEDIUMBLOB is a blob column with a maximum length of
16777215 which can be mapped to DB2 BLOB(16M).
– MEDIUMTEXT
MySQL MEDIUMTEXT is a text column with a maximum length of
16777215 which can be mapped to DB2 CLOB(16M).
– LONGBLOB
MySQL LONGBLOB is very large BLOB column with a maximum length of
4294967295 which can be mapped to DB2 BLOB(2G).
– LONGTEXT
MySQL LONGTEXT is text column with a maximum length of 4294967295
which can be mapped to DB2 CLOB(2G).
– ENUM
MySQL has a special ENUM type, which is a string object can have only
one value chosen from a list of values 'value1', 'value2',..., NULLI which
can be mapped to DB2 VARCHAR() with check constraints.
– SET
MySQL has another special SET type which is a string object that can
have zero or more values, which must be chosen from the list of values
'value1', 'value2',... which can be mapped to DB2 VARCHAR() with check
constraints.

Table 6-1 shows the default data type mappings between the two databases
used by the IBM Data Movement Tool. This is the mapping we used for our
sample conversion.

120 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Table 6-1 MySQL to DB2 data type mapping

MySQL 5.1 DB2 9.7
TINYINT SMALLINT

TINYINT SMALLINT
UNSIGNED

SMALLINT SMALLINT
SMALLINT INTEGER
UNSIGNED Optional: SMALLINT

BIT SMALLINT

BOOLEAN SMALLINT

MEDIUMINT INTEGER

MEDIUMINT INTEGER
UNSIGNED

INTEGER/INT INTEGER

INTEGER/INT BIGINT
UNSIGNED Optional: INTEGER
BIGINT BIGINT

BIGINT DECIMAL
UNSIGNED Optional: BIGINT
FLOAT DOUBLE

FLOAT DOUBLE
UNSIGNED
DOUBLE DOUBLE

DOUBLE DECIMAL
UNSIGNED Optional: DOUBLE
REAL DOUBLE

REAL DOUBLE
UNSIGNED

NUMBERIC DECIMAL(31,0)
DECIMAL
DEC
NUMERIC(P) DECIMAL(min(P,31),0)
NUMERIC(P,0)
DECIMAL(P)
DECIMAL(P,0)
DEC(P)
DEC(P,0)

DECIMAL DECIMAL
UNSIGNED

Chapter 6. Database conversion 121

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

NUMERIC DECIMAL
UNSIGNED

NUMBERIC(p,s) DECIMAL(min(p,32), min(s,32))

DECIMAL(p,s) DECIMAL(min(s,32), min(s,32))
DEC(p,s) DECIMAL(min(p,32),0)
where:
s > 0 && p >= s
s > 0 && p < s
s<0

DATE DATE

DATETIME TIMESTAMP
Optional: TIME

TIMESTAMP TIMESTAMP

TIME TIME

YEAR CHAR(4)
Optional: SMALLINT

CHAR CHAR

VARCHAR VARCHAR
BINARY CHAR(I) FOR BIT DATA

VARBINARY VARCHAR(I) FOR BIT DATA

TINYBLOB BLOB(255)
Optional: VARCHAR(255)

TINYTEXT CLOB(255)

BLOB BLOB(65535)

TEXT CLOB(65535)

MEDIUMBLOB BLOB(16777215)

MEDIUMTEXT CLOB(16777215)

LONGBLOB BLOB(2G)

LONGTEXT CLOB(2G)

ENUM VARCHAR() with check constraints

SET VARCHAR() with check constraints

6.2 Data definition language differences

In this section we address the Data Definition Language (DDL) syntax
differences between MySQL and DB2 DDL statements. These differences can
be syntactical, semantic, or functional.

122 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Both MySQL and DB2 follow the structured query language (SQL), standardized
language used to access databases and their objects, as defined by the
ANSI/ISO.

The data definition language is a set of SQL statements. These can be used for a
variety of tasks including, creation or deletion of databases and database objects
(tables, views, and indexes), definitions of column types, and definitions of
referential integrity rules.

6.2.1 Database manipulation

By default, MySQL database objects are stored in a single directory. Although
this is simple, it poses a big performance bottleneck due to slow disk seek, slow
search, and non-indexing of data. MySQL depends on the operating system's
capabilities for distributing its data across disks. It uses symbolic links to link
different disks for different databases, as well as for database and table
distribution.

On Linux machines, MySQL can use the file system mounting options, however,
in most cases, MySQL uses symbolic links. This can be done by creating a
directory where you have extra space:
bash> cd <file system with space>
bash> mkdir mysqldata

then creating a symbolic link to the newly created directory using:

bash> cd /var/lib/mysql
bash>ln -sf <file system with space>/mysqldata data

then creating a database from the mysql prompt using:

mysql>CREATE DATABASE inventory

MySQL users can distribute tables using symbolic linking or the data and index
directory options of the CREATE TABLE statement.

MySQL stores data in single files, multiple files, or table spaces, depending on
the table type being used. Figure 6-3 shows example of storage engines that fall
under one of these three types.

Chapter 6. Database conversion 123

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

Figure 6-3 MySQL physical storage levels

The tables to the left side of the diagram are managed by the MyISAM storage
engine. For MyISAM tables, MySQL creates a .MYD file for data and a .MYI file
to store indexes, with only one file for all data and indexes, respectively. The
tables in the middle of the diagram are managed by the Merge storage engine.
With Merge tables, the .MRG file contains the names of the tables that should be
used and a .FRM file for table definition. In a Merge table various tables are
used, each of these having there own data file. However, as a whole, a Merge
table uses multiple data files. The tables to the right side of the diagram are
managed by InnoDB storage engine. For InnoDB tables, MySQL stores data in a
table space identified by the path parameter, innodb_data_file_path. Multiple
data files can be used for InnonDB.

MySQL also has a feature called user-defined partitioning, which allows for table
data to be horizontally split across file systems depending on a specific set of
data defined by the user. For each partition created there is a corresponding
.MYD file for data and .MYI file for the index.
In contrast to MySQL, DB2 stores everything in table spaces. Table spaces are
logical representations of physical containers on the file system. DB2 uses a
better approach for the logical and physical distribution of the database and the
database elements in different sectors, as shown in Figure 6-4. After completing
the conversion of your database from MySQL to DB2, you can use these features
to enhance the performance of your application.

124 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Figure 6-4 DB2 storage levels

Instance
A DB2 server can have more than one instance. One instance can have multiple
databases. One instance per application database has the advantage that the
application and database support do not have to coordinate with one another to
take the database or the instance offline. For conversion purposes, a single
instance can be created for your database application environment using the
db2icrt command:
bash> db2icrt -u db2fenc1 db2inst1

Database
A database represents your data as a collection of data in structured fashion. It
includes a set of system catalog tables that describe the logical and physical
structure of the data, a configuration file containing environment parameter
values used by the database, and a recovery log with ongoing and archivable
transactions.

Database Partition
A database partition is part of a database containing its own data, indexes,
configuration files, and transaction logs. A database partition is sometimes called
a node or a database node. A partitioned database environment is a database
installation that supports the distribution of data across database partitions. This

Chapter 6. Database conversion 125

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

can be used if you want to spread your DB2 database across multiple servers in
a cluster or along multiple nodes. There are no database partition group design
considerations when using a non-partitioned database. The database partition
group can be created within a database using the following command:
db2> CREATE DATABASE PARTITION GROUP MaxGroup ON ALL DBPARTITIONNUMS

Creating a database
The database in DB2 can be created simply by issuing the following command:
db2>CREATE DATABASE invent

This command generates a new database with a default path, and table spaces.
It creates three initial table spaces and the system tables, and creates the
recovery log.

You can use the create database statement with options to personalize the
database and take advantage of DB2 advanced features such as automatic
storage which simplifies storage management for table spaces as shown in
Example 6-1. When using automatic storage, it is possible to specify a group of
storage devices for DB2 to use for your database. This allows DB2 to allocate
and grow this specified space as table spaces are created and populated.
Automatic storage is turned on by default when creating a database.

Example 6-1 Create database

db2> CREATE DATABASE invent AUTOMATICE STORAGE YES on '/db2fs/invent'

Dropping a database
In MySQL you can drop the database using:
mysql> DROP DATABASE [if exists] inventory

This removes all the database files (.BAK, .DAT, .HSH, .ISD, .ISM, .ISM, .MRG,
.MYD,.MYI, .db, .frm) from your file system.

DB2 has a similar command:

db2> DROP DATABASE invent [at DBPARTITIONNUM]

This command deletes the database contents and all log files for the database,
uncatalogs the database, and deletes the database subdirectory.

Alter database
MySQL alter database command allows to change the overall characteristics of a
database. For example, the character set clause changes the database

126 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

character set, and the collation clause changes the database collation. The basic
syntax for altering the database is:
mysql>ALTER DATABASE inventory CHARACTER SET charset_name COLLATE
collation_name

In DB2 you can use the UPDATE DATABASE CONFIGURATION and UPDATE
DATABASE MANAGER CONFIGURATION commands to set the database and
database manager configuration parameters. These allow modification of various
configuration parameters such as log file size, log file path, heap size, cache
size, and many others. Advantage can be taken of the DB2 autonomic features
by enabling the automatic maintenance and self tuning memory manager
features. Automatic maintenance allows for scheduling database backups,
keeping statistic current, and reorganizing tables and indexes. Self tuning
memory manager provides constant tuning of your database without the need for
DBA intervention. The following are examples of how these can be set in DB2:
db2> UPDATE DATABASE MANAGER CONFIGURATION using diaglevel 3
db2> UPDATE DATABASE ONFIGURATION for invent
using auto_maint on
auto_tbl_maint on
auto_runstats on
auto_reorg on
self_tuning_mem on

The example above does not show all the parameters available in DB2. For more
information on how to setup automatic maintenance and self tuning memory
manager, visit at
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

In addition, these commands can be used to change the physical and logical
partitioning of a database, to allocate table spaces and paging configurations.

Table space
A table space is a storage structure containing tables, indexes, large objects, and
long data. Table spaces reside in database partition groups and allow
assignment of database location and table data directly onto containers. DB2
allows for two types of table spaces: System Managed Space (SMS), where the
operating system allocates and manages the space where the tables are stored,
and Data Managed Space (DMS), where the database administrator has the
ability to decide which devices or files to use and allows DB2 to manage this
space. Another option is to enable automatic storage for the table spaces. No
container definitions are needed in the latter case because the DB2 database
manager assigns and manages the container automatically.

Any DB2 database should at least have the following three table spaces:

Chapter 6. Database conversion 127

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 One catalog table space, which contains system catalog tables

򐂰 One or more user table space, which contains user defined tables
򐂰 One or more temporary table space, which contains temporary tables

More table spaces can be created using the following commands:

db2> CREATE REGULAR TABLESPACE tblsp1 PAGESIZE 8192 MANAGED BY SYSTEM using
('/home/db2inst1/database/user8K') extentsize 8 prefetchsize 8 bufferpool
bp8k
db2> CREATE TABLESPACE tblsp2 MANAGED BY DATABASE using (device
'/dev/rhdisk6' 10000, device '/dev/rhdisk7' 10000, device '/dev/rhdisk8'
10000) overhead 12.67 transferrate 0.18
db2> CREATE TABLESPACE tblsp3 MANAGED BY AUTOMATIC STORAGE

Schema
A schema is an identifier such as a user ID that helps group tables and other
database objects. A schema can be owned by an individual, and the owner can
control access to the data and the objects within it. A schema is also an object in
the database. It may be created automatically when the first object in a schema
is created. We can create a schema using:
db2>CREATE SCHEMA inventschema AUTHORIZATION inventUser

6.2.2 Table manipulation

Tables are logical structures made up of columns and rows which are maintained
by the database manager.

MySQL tables
As shown in Figure 6-5 on page 129, MySQL supports two types of tables:
transaction-safe tables and non transaction-safe tables. Transaction-safe tables
(managed by InnoDB or NDB storage engines) are crash safe and can take part
in transactions providing concurrency features that allow commit and rollback.
On the other hand, non transaction-safe tables (managed by MyISAM,
MEMORY, MERGE, ARCHIVE or CSV storage engines) are less safe but are
much faster, and consume less space and memory.

128 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Figure 6-5 MySQl table types

The following tables are the basic storage elements in MySQL:

򐂰 InnoDB table
InnoDB is a transaction safe storage engine with commit, rollback, and crash
recovery capabilities. It provides locking on row level and consistent
non-locking read-in select statements.
InnoDB is a complete database service placed under MySQL having its own
buffer pool for caching data and indexes in main memory and supporting
multiple table spaces. An InnoDB table can be created using the following
command:
mysql>CREATE TABLE tblinnodb (i int ,f float ) type=innodb;
򐂰 NDB
NDB is a transaction safe storage engine. This storage engine allows for
clustering of a database in a shared-nothing architecture. This type of
architecture allows tables to be split across multiple servers, with each server
having its own set of data. In order for the table to be a part of the cluster, the
table must be using NDB storage engine.
򐂰 MyISAM table
MyISAM is the default table type in MySQL. It is based on the ISAM code,
and supports concurrent insert, big files on operating system and file system
that support big files, better indexing, and index compression and table
compression.
MyISAM is a default MySQL table so it can be created either with specifying
type=myisam or not specifying any type:
mysql>CREATE TABLE tblmyisam (i int ,f float );
򐂰 MERGE

Chapter 6. Database conversion 129

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

MERGE storage engine is a group of MyISAM tables in one table, grouped

together across the same disk or multiple disks. The tables must have exact
column and index structures.
򐂰 MEMORY (HEAP)
MEMORY was previously called the HEAP storage engine. Tables that are
created in the MEMORY storage engine are stored in memory. The table
definition is stored on disk but the rows are stored in memory. Therefore, the
table still exists after a reboot, however, all rows within this table are lost.
򐂰 ARCHIVE
ARCHIVE stores that data in compressed format on disk and is usually used
to store large amounts of data. There are no indexes in the ARCHIVE storage
engine.
򐂰 CSV
The CVS storage engine which stands for 'Comma Separated Value' allows
data to be stored in text files. This data can then be accessed via SQL calls.
򐂰 FEDERATED
FEDERATED storage engine work with non transactional tables and allows
tables to be accessed on a remote server as if they were stored locally. Table
definition is stored locally, however, no data is actually stored on the local
server.

DB2 conversion of MySQL table

In DB2 all the tables take part in the transaction. Therefore, tables that are
managed by MySQL InnonDB, MyISAM, ARCHIVE, and CSV storage engines
can all be converted to a DB2 regular table. A DB2 regular table is a general
purpose table which is created with the CREATE TABLE statement and is used
to hold persistent user data.

Create table syntax

In this section we give a high level overview of the difference in the CREATE
TABLE syntax of MySQL and DB2. MySQL CREATE TABLE statements are
quite simple with few exceptions. Example 6-2 on page 130 shows the creating
of MySQL InnoDB. Example 6-3 shows the DB2 conversion. You can notice that
no major change is required.

Example 6-2 Creating MySQL InnoDB managed table

mysql>CREATE TABLE tblinno(col1 int, col2 char(10)) engine=InnoDB ;

130 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Example 6-3 DB2 conversion of creating MySQL InnoDB managed table

db2>CREATE TABLE tblinno(col1 int, col2 char(10))

Example 6-4 shows how to create a table using the MyISAM storage engine.
Example 6-5 is the DB2 conversion. Notable changes are:
򐂰 Changes in the data type according to data type mapping.
򐂰 Instead of auto_increment, generated by default as identity is used.

Example 6-4 Creating MySQL MyiSAM table

mysql>CREATE TABLE tblmysiam (
wk_id int(11) unsigned NOT NULL auto_increment,
user_id int(11) unsigned default NULL,
cnt int(10) unsigned default NULL,
cat_id int(12) unsigned default NULL,
status varchar(10) default NULL,
PRIMARY KEY (wk_id)) type=MyISAM;

Example 6-5 Conversion of MySQL MyISAM table creation

db2>CREATE TABLE tblmysiam (
wk_id INT NOT NULL GENERATED BY DEFAULT AS IDENTITY,
user_id INT default NULL,
cnt INT default NULL,
cat_id INT default NULL,
status VARCHAR(10) default NULL,
PRIMARY KEY (wk_id));

Example 6-6 is a MySQL table creation example using the ARCHIVE engine.
Example 6-7 is the DB2 conversion using row compression.

Example 6-6 Creating MySQL ARCHIVE table

mysql>CREATE TABLE tblarchive(col1 int, col2 char(10)) type=ARCHIVE;

Example 6-7 DB2 conversion of creating MySQL ARCHIVE table

db2>CREATE TABLE tblarchive(col1 int, col2 char(10)) COMPRESS YES

Example 6-8 shows how to create a MySQL table with Partitioning. Example 6-9
is the DB2 conversion, again no major change required.

Example 6-8 Creating MySQL table using partitioning with the default MyISAM storage
engine
mysql>CREATE TABLE partsales (id INT, item VARCHAR (20) )
PARTITION BY RANGE (id)(

Chapter 6. Database conversion 131

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

PARTITION p1 VALUES LESS THAN (10),

PARTITION p2 VALUES LESS THAN (20)
);

Example 6-9 DB2 conversion of MySQL table creation with partitioning

db2>CREATE TABLE partsales (id INT, item VARCHAR (20)
PARTITION BY RANGE (id)(
PARTITION p1 STARTING FROM (MINVALUE) ENDING (10),
PARTITION p2 ENDING (20)
);

Alter table
Alter table is a statement used to modify one or more properties of a table. The
syntax of the ALTER TABLE statement for MySQL and DB2 is quite similar and
is shown in Example 6-10.

Example 6-10 MySQL and DB2 alter table example

mysql>ALTER TABLE partsales modify status varchar(20);

db2>ALTER TABLE partsales alter column status set data type varchar(20)

Alter table in DB2 now supports the dropping of columns. DB2 dropping column
using temporary table. The syntax of ALTER TABLE for MySQL and DB2 is
similar as shown in Example 6-11.

Example 6-11 MySQL and DB2 dropping column example

mysql> ALTER TABLE sales drop column c1, drop column c2;

db2> ALTER TABLE sales drop c1 drop c2

Drop table
Tables can easily be deleted from the database by issuing the DROP TABLE
statement as shown below.

For MySQL:
DROP [TEMPORARY] TABLE [IF EXISTS] tbl_name [, tbl_name,...] [RESTRICT
|CASCADE]

For DB2:
DROP table tbname

132 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Non-persistent storage tables

The following discusses the difference between MySQL and DB2 temporary
table types.
򐂰 MERGE table
A merge table groups multiple MyISAM tables into one table. Dropping the
merge table leads to dropping of the merge specifications but tables that
made up the merge table continue to exist.
Tables can be merged only if they have identical columns and key
information. Example 6-12 shows the statements for merge table in MySQL.

Example 6-12 Usage of merge table in MySQL

mysql> CREATE TABLE table1
(col1 int not null auto_increament primary key, col2 char(20));
mysql> CREATE TABLE table2
(col1 int not null auto_increament primary key,col2 char(20));
mysql> CREATE TABLE tblmerge
(col1 int not null auto_increament, col2 char(20), key(col1));
TYPE=MERGE UNION=(table1,table2) INSERT_METHOD=LAST;
mysql> INSERT INTO table1 (col2) VALUES ("value1"),("value2");
mysql> INSERT INTOtable2 (col2) VALUES ("value3"),("value4");

DB2 uses the updateable UNION ALL view to achieve the above feature.
UNION ALL views are commonly used for logically combining different but
semantically related tables. UNION ALL view is also used for unification of
like tables for better performance, manageability, and integrating federated
data sources:
An example of using the UNION ALL command for views is:
db2>CREAT VIEW UNIONVIEW as SELECT * FROM table1 UNION ALL SELECT * FROM
table2
򐂰 MEMORY table
A MEMORY table is a hashed index that is always stored in memory. Memory
tables are fast but not crash safe. When MySQL crashes or has a scheduled
reboot the MEMORY table will still exists on reboot but data is lost.
A MySQL Memory table can be created using the following command:
mysql> CREATE TABLE memtable type=MEMORY SELECT * FROM table1;
MySQL MEMORY tables can be converted to DB2 as temporary tables,
materialized query tables, or indexes depending upon your requirements.
– Temporary table
DB2 temporary tables are tables used for storing data in non-persistent,
in-memory, session-specific tables. Once a session is over, the table

Chapter 6. Database conversion 133

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

definition for the table is lost. When your application is using MEMORY
table in this fashion, temporary tables can be declared in your application
by calling the statements:
• Create user temporary table space if it does not exist using:
db2>CREATE USER TEMPORARY TABLESPACE discompose MANAGED BY SYSTEM
using ('usertemp1')
• Declare temporary table in the application:
db2>DECLARE GLOBAL TEMPORARY TABLE distemper LIKE table1 ON COMMIT
db2>DELETE ROWS NOT LOGGED IN discompose
– Materialized query table
Materialized query tables (MQT) also known as summary table can also
be used to improve the query performance. A MQT is a table whose
definition is based on the results of a query, and whose data is in the form
of pre-computed results. If the SQL compiler determines that a query will
run more efficiently against a materialized query table than a base table or
tables, the query executes against the materialized query table:
db2>CREATE TABLE sales AS (SELECT * FROM table1) DATA INITIALLY
DEFERRED REFRESH DEFERRED
We discuss MQT in more detail in 11.4, “Materialized query tables (MQT)”
on page 391.
In addition, DB2 also supports tables for clustering and query performance
enhancement. These tables can also be used according to various
requirements:
򐂰 Multidimensional clustering (MDC) tables
Multidimensional clustering (MDC) tables have a physical cluster on more
than one key or dimension at the same time. An MDC table maintains
clustering over all dimensions automatically and continuously, thus
eliminating the need to reorganize the table in order to restore the physical
order of the data.
When creating MDC tables, the performance of many queries might improve
because the optimizer can apply additional optimization strategies. Some of
the advantages of MDC tables include quicker and more rare scanning
because of dimension block, faster lookups, block level index ANDing and
ORing, and faster retrieval.
The following command can be issued to create it:
db2>CREATE TABLE tblmdc
(col1 int, col2 int, col3 int, col4 char(10))
ORGANIZE BY DIMENSIONS(col1,col2,col3)

134 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

We discuss MQT in more detail in 11.3.3, “Multidimensional clustering” on

page 389.
򐂰 Range-clustered tables (RCT)
RCT are implemented as sequential clusters of data that provide fast, direct
access. It is a table layout scheme where each record in the table has a
predetermined offset from the logical start of the table. Some advantages
associated with RCT are direct and hence quicker access times, less
maintenance, less logging, less locking, less buffer pool size, and less
indexing. It can be created using:
db2> CREATE TABLE tblrct
(col1 int not null, col2 int not null,col3 char(10),
col4 float)
ORAGANIZED BY KEY SEQUENCE
(col1 starting from 1 ending at 10000)
allow overflow
򐂰 Typed table/hierarchy table
Typed tables are tables that are defined with a user-defined structured type.
With typed tables you can establish a hierarchical structure with a defined
relationship between tables called a table hierarchy. The table hierarchy is
made up of a single root table, supertables, and subtables. It can be created
as shown in Example 6-13

Example 6-13 Usage of typed/hierarchy table

//Here is the SQL to create the BusinessUnit typed table

db2>CREATE TABLE BusinessUnit OF BusinessUnit_t (REF IS Oid USER GENERATED)

//Here is the SQL to create the tables in the Person table hierarchy

db2>CREATE TABLE Person OF Person_t (REF IS Oid USER GENERATED)

db2>CREATE TABLE Employee OF Employee_t UNDER Person INHERIT SELECT PRIVILEGES
(SerialNum WITH OPTIONS NOT NULL, Dept WITH OPTIONS SCOPE BusinessUnit )
db2> CREATE TABLE Student OF Student_t UNDER Person INHERIT SELECT PRIVILEGES
db2> CREATE TABLE Manager OF Manager_t UNDER Employee INHERIT SELECT PRIVILEGES
db2> CREATE TABLE Architect OF Architect_t UNDER Employee INHERIT SELECT
PRIVILEGES

򐂰 Views
Views are the named specification of a result table. This specification is a
select statement that is run whenever the view is referenced in an SQL
statement. It can be used just like a base table.
A simple view can be created by issuing a create statement as shown in
Example 6-14

Chapter 6. Database conversion 135

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

Example 6-14 Create view example

db2>CREATE TABLE table1(col1 INTINT, col2 INT, col3 CHAR(20),col4 FLOAT,
col5 CHAR(30))
db2>CREATE TABLE itable2(col6 INT, col7 INT, col8 CHAR(20),col9 FLOAT,
col10 CHAR(30))
db2>CREATE VIEW myview(col1,sum1,col4,col10) AS
SELECT col1,col1+col6,col4,col10
FROM table1, table2
WHERE col1<10 AND col6>10

6.2.3 Index manipulation

An index is an object in the database system, which uses techniques for faster
retrieval of the data from tables. It is an ordered set of pointers to rows of a base
table managed by the database manager directly.

MySQL supports both single column indexes and multi-column indexes. MySQL
has five types of indexes:
򐂰 Primary key
򐂰 Unique
򐂰 Non-unique
򐂰 Fulltext
򐂰 Spatial

DB2 supports all the index types supported by MySQL with the same kind of
terminologies allowing them to map directly during conversion.

Create Index
The following is the MySQL CREATE INDEX syntax:
CREATE [ONLINE|OFFLINE] [UNIQUE|FULLTEXT|SPATIAL] INDEX index_name
[index_type]
ON tbl_name (index_col_name,...)
index_col_name:
col_name [(length)] [ASC | DESC]

Creating an index in DB2 is quite similar:

CREATE [unique] INDEX index-name on tablename (columnnames ASC|DESC)
SPECIFICATION ONLY INCLUDE(column-name)
CLUSTER/EXTEND USING index-extension-name (constant-expression)
PCTFREE 10/PCTFREE integer LEVEL PCTFREE integer MINPCTUSED integer
ALLOW/DISALLOW REVERSE SCANS
PAGE SPLIT SYMMETRIC/PAGE SPLIT HIGH/LOW
COLLECT STATISTICS DETAILED SAMPLED

136 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Drop index
Drop index statement for MySQL and DB2:
mysql> DROP INDEX index_name ON tbl_name;
db2> DROP INDEX index_name

6.2.4 Triggers manipulation

Triggers are a predefined set of actions to be executed automatically in response
to a set of events. Triggers are commonly defined to respond to the following
events, INSERT, UPDATE, or DELETE.

There are two types of triggers - statement and row triggers. Statement triggers
are executed in response to a single INSERT, UPDATE, or DELETE statement.
Row triggers are executed for each row affected by the INSERT, UPDATE, or
DELETE statement. There are many benefits to having triggers, since they allow
logging of information about changes to a table, can be used to validate an
insert, restrict access to specific data, or make data modifications and
comparisons at change.

DB2 supports both row based and statement based triggers; MySQL supports
only row-based triggers. Both MySQL and DB2 support INSERT, UPDATE, and
DELETE triggered events and can be defined to fire using BEFORE and AFTER.
DB2 has an additional firing classification called INSTEAD OF which allows the
triggers to perform INSERT, UPDATE, or DELETE on views. MySQL does not
support a trigger on a view.

The following is the create trigger syntax for MySQL:

CREATE [DEFINER = { user | CURRENT_USER}]
TRIGGER trigger_name trigger_time trigger_event
ON tbl_name FOR EACH ROW trigger_stmt

Creating a trigger in DB2 has similar syntax:

CREATE TRIGGER trigger-name NO CASCADE BEFORE/AFTER/INSTEAD OF
INSERT/DELETE/UPDATE OF column-name ON table-name/view-name
REFERENCE OLD AS correlation-name/NEW AS correlation-name/OLD TABLE
AS identifier/NEW TABLE AS identifier
FOR EACH ROW/FOR EACH STATEMENT triggered-action

6.2.5 Procedures and function manipulation

Stored procedures and functions are routines that are stored and run on the
database server. Stored procedures can be developed to compute or manipulate
results from a query on the database server prior to sending it back to the

Chapter 6. Database conversion 137

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

application. This can result in increased performance since unnecessary data is

not sent across the network to the application. This also decreases dependency
between the application and the database since the application no longer needs
to manipulate the results. Both DB2 and MySQL follow the standard syntax
specified by the SQL:2003 for stored procedures and functions, therefore,
creating a procedure or function in DB2 is similar.

6.3 Other considerations

Up until now we discussed approaches for converting database elements, which
use similar approaches in the two products. In subsequent sections we discuss
the conversion of database objects that do not map directly from MySQL to DB2.
We also discuss server and database placement architecture.

Multiple servers
In some cases multiple MySQL servers are placed on the same machine.
Possible cases reasons include user management, testing, or potentially
differentiating applications. MySQL provides an option to run multiple servers on
same machine using several operating parameters.

There are several ways to configure a new server; we used:

bash> /path/to/mysqld --socket=file_name --port=port_number

or you can use:

bash> MYSQL_UNIX_PORT=/tmp/mysqld-new.sock
bash> MYSQL_TCP_PORT=3307
bash> export MYSQL_UNIX_PORT MYSQL_TCP_PORT
ash> mysql

DB2 supports similar functionality using multiple database manager instances on

the same machine. Each having its own configuration files, directories, and
database. Figure 6-6 shows the typical scenario.

DB2 also supports the creation of different instances on the same machine with a
different DB2 version.

138 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Figure 6-6 Typical DB2 setup

Multiple databases and schema conversion

Within MySQL, tables in a database cannot be logically grouped by a schema or
qualified by a name based on the application. To fulfill this requirement MySQL
tables are placed in separate database.

MySQL supports access of tables in multiple databases using the same

connection. So an application connected to MySQL can use two tables in
different databases in a single statement. The queries in Example 6-15 shows
how it works

Example 6-15 Multiple database example

mysql>CREATE DATABASE mydb1;
mysql>CREATE DATABASE mydb2;
mysql>CONNECT mydb1;
mysql>CREATE TABLE table1(col1 INT, col2 CHAR(10));
mysql>INSERT INTO table1 VALUES(1,"new");
mysql>CONNECT mydb2;
mysql>CREATE TABLE table2(col1 INT, col2 CHAR(10));
mysql>INSERT INTO table2 VALUES(1,"val 1");
mysql>SELECT * FROM table2, mydb1.table1;

This works in MySQL and gives the following result

+------+-------+------+------+
| col1 | col2 | col1 | col2 |

Chapter 6. Database conversion 139

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

+------+-------+------+------+
| 1 | val 1 | 1 | new |
| 1 | val 1 | 2 | new |
+------+-------+------+------+

Figure 6-7 shows the architecture of multiple MySQL databases accessed by a

single application using the same connection.

Figure 6-7 MySQL application using multiple DBs instead of multiple schema

DB2 supports access to multiple databases using database links in a federated

system. For a non-federated system, DB2 applications use a more logical
technique by using multiple schemas to replace use of multiple databases within
the same application. Each database can have multiple schemas and each table
belongs to a particular schema.

When converting the MySQL applications using multiple databases, all

databases used in the applications can be placed under a single DB2 database.

Each MySQL database becomes represented by a DB2 schema to hold the

tables under that database. All tables can be accessed by the application using
single connection. The DBA only needs to manage one database.

DB2 schema can be created using the following command:

db2>CREATE SCHEMA schema1 AUTHORIZATION schema1

The tables created in the particular schema can be accessed using the full table
qualifier table schema.table name.

Table placement
MySQL does not support table spaces for managing physical location and page
size or distributing tables onto the different table spaces, except with the optional
InnoDB storage engine which supports multiple table spaces by distributing them
into different files.

140 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

DB2 supports table spaces to establish the relationship between the physical
storage devices used by your database system and the logical containers used
to store data. Table spaces reside in database partition groups. They allow you
to assign the location of table data directly onto containers.

Prior to conversion of your database structure, proper table spaces of different

sizes should be created in DB2. Individual or multiple tables can then be
assigned to the tables spaces. If the table space design is done effectively it can
greatly increase performance. We discuss this topic in more detail in Chapter 10,
“Testing and tuning” on page 321.

List information
MySQL provides a show command to list the information about databases,
tables, columns, or status information about the server.

DB2 provide commands for getting information about instances, databases, table
spaces, and others. The DB2 system catalogues contain all necessary
information about tables, columns, and indexes and others. You can use
describe and list commands to display database and table structure or use
select statement to get the details of the table definition.

Table 6-2 shows examples of MySQL and corresponding DB2 statements or

commends to list database or table related information.

Table 6-2 MySQL to DB2 conversion of list information statement

MySQL DB2

show databases list database directory

show tables from <databasename> list tables

show columns from <tablename> describe table <tablename>

show index from <tablename> SELECT indname FROM syscat.indexes

WHERE tabname=tablename

Referential integrity refers to the constraints defined on the table and its columns,
which help you to control the relationship of data in different tables. Essentially,
this involves primary keys, foreign keys, and unique keys.

Primary keys and unique keys are treated similarly in MySQL and DB2.
However, MySQL currently only parses foreign key syntax in the CREATE
TABLE statements, but does not use/store the information about foreign keys
except in InnoDB tables which support checking of foreign key constraints,
including CASCADE, ON DELETE, and ON UPDATE.

Chapter 6. Database conversion 141

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

DB2 provides full support for foreign keys. With the full referential integrity
functionality from DB2, your application can be released from the job of taking
care of the data integrity issues. Example 6-16 shows creation and usage of
foreign key constraints in DB2.

Example 6-16 Foreign key constraint usage

db2> ALTER TABLE table1 ADD CONSTRAINT foreign1 FOREIGN KEY (id)
REFERENCES table2 ON DELETE SET NULL

We discuss more foreign keys creation in more detail in 6.5.2, “Manual database
object conversion and enhancements” on page 158.

6.4 Converting database

After understanding MySQL database structure/schema relative to different
types of DB2 objects, such as database, tables, views, indexes and referential
integrity, this section provides details of the database structure/schema
conversion from existing MySQL database to DB2.

Database schema conversion can be done in various ways, however, the most
common approaches are:
򐂰 Automatic conversion using porting tools
򐂰 Manual porting
򐂰 Metadata transport

In general, all approaches taken above use existing MySQL databases as input
and pass them through the following functional engine:
򐂰 Capture database schema information from MySQL
򐂰 Modify schema information for DB2
򐂰 Create the database in DB2 with structure

6.4.1 Automatic conversion using porting tools

Using porting tools is the most common and easiest approach. The IBM Data
Movement Tool can simplify your move to DB2. There are a number of third party
upgrade tools available on the market which can be used but IBM does not
guarantee the correct functionality of these tools.

IBM Data Movement Tool

The IBM Data Movement Tool is a simple conversion tool that takes the MySQL
database to be converted, and creates the database objects and data records for

142 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

the new DB2 database. With the IBM Data Movement Tool, you can
automatically convert data types, tables, columns, and indexes into equivalent
DB2 database elements. The IBM Data Movement Tool provides database
administrators (DBAs) and application programmers the functionality needed to
automate the conversion task. The strength of this tool is shown in large scale
data movement projects. This tool has been used to move up to 4TB of data in
just three days with good planning and procedures.

You can reduce the downtime, eliminate human error, and cut back on person
hours and other resources associated with the database conversion by using the
following features found in the IBM Data Movement Tool:
򐂰 Extract of DDL statements from the MySQL source database.
򐂰 Extract data from the MySQL source database.
򐂰 Generate and run DB2 DDL conversion scripts.
򐂰 Automate the conversion of database object definitions.
򐂰 View and refine conversion scripts.
򐂰 Efficiently implement converted objects using the deployment options
(interactive deployment or automated deployment).
򐂰 Generate and run data movement scripts.
򐂰 Track the status of object conversions and data movement, including error
messages, error location, and DDL change reports, using the detailed
conversion log file and report.

Instructions on download, installation, and usage for the IBM Data Movement
Tool are described in 5.3, “IBM Data Movement Tool installation and usage” on
page 111.

For the rest of the chapter we discuss how to convert the database structure to
DB2. The next chapter discusses how to convert the data to DB2. The following
briefly describes the database structure conversion process when using the IBM
Data Movement Tool:
1. Specify source and DB2 database server connection information.
2. Test the connection to the source and target database. Click Connect to
MySQL to test the connection and Connect to DB2 to test the DB2
connection.
3. Specify the working directory where DDL and data are to be extracted to.
4. With the IBM Data Movement Tool you have the option to extract only the
database objects, the data or both. Choose if you want DDL and/or DATA.

Chapter 6. Database conversion 143

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

5. Click the Extract Data button to extract the DDL and/or DATA and
automatically convert to DB2 syntax. You can monitor progress in the console
window.
6. After the data extraction is completed successfully, go through the result
output files for the status of the data movement, warnings, errors, and other
potential issues. Optionally, you can click on the View Script/Output to
check the generated scripts, DDL, data or the output log file.
7. Click the Deploy Data button to automatically create tables, indexes in DB2
and load data that was extracted from the source database. Optionally, you
can click Interactive Deploy to deploy database objects one by one.

In this chapter we only discuss the database conversion portion of the

conversion process.

Other upgrade tools

There are other upgrade tools available to port MySQL database to DB2.
򐂰 SQLWays
SQLWays is a database conversion tool from Ispirer Systems Ltd.

6.4.2 Manual conversion

In the previous section we discuss conversion using the IBM Data Movement
Tool. Sometime it is required to follow more personalized manual process to
convert your database instead of the standard tool. For those few cases we
describe manual conversion processes, with a focus on converting MySQL
objects and features which are not automatically converted by the IBM Data
Movement Tool.

In 6.1, “Data type mapping” on page 116 and 6.2, “Data definition language
differences” on page 122, we demonstrated syntax and semantic differences
between MySQL and DB2, which could use manual conversion. We have also
discussed creation, deletion, and alteration of various database objects such as
database, tables, index, views among others, and how these are related when
converting from MySQL and DB2. There are several steps involved in completing
the manual process, which will be discussed in more detail below:
򐂰 Capturing the database schema information from MySQL
MySQL offers a utility called mysqldump which extracts the database
structure and deposits this information into a text file. The structure is
represented in DDL, and can be used to recreate these database elements
for your DB2 server. Syntax for mysqldump is as shown below:
bash> mysqldump DatabaseName > mysqlobjects.ddl

144 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

When using this tool on a very large database it is recommended to use the
--quick or --opt option. Without these options, mysqldump loads whole result
set into memory before dumping the results.
For further information on this utility refer to the MySQL Reference Manual at:
http://dev.mysql.com/doc/refman/5.1/en/
򐂰 Modify schema information for DB2
Now that you have captured the source of your MySQL database structure
using the mysqldump utility, it is time to modify the schema information and
make it work for DB2. Manual changes include:
– DDL changes
The first step in schema modification is the conversion of the create
statement for various database objects such as database, tables, views,
indexes and others. Refer to 6.2, “Data definition language differences” on
page 122 or these conversions as reference. Also a DDL script must be
written for creating the database and table spaces.
– Data type changes
Check the data types used in the table definition. Change the MySQL data
type to DB2 data types. Refer 6.1, “Data type mapping” on page 116.
– Reserved words conversion
There are many reserved words in MySQL and DB2 which cannot be a
valid name for the column and database element. Refer to the MySQL
Reference Manual (http://dev.mysql.com/doc/refman/5.1/en/) and DB2.
For more detail about reserved words and change the conflicting names in
the DDL statements, visit DB2 Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp
– Create database and database objects
Now that you have DDL statements modified, the DB2 database and
database objects themselves need to be created. This can be done from
the command line processor (CLP).

It is a common practice to place the database and table space creation

statements in one file. Database object creation such as tables, views, and
others should be placed in a separate file. Note that database and table spaces
must be created prior to creating the database objects. The database creation
scripts should be based on the logical and physical database design. In our
example, we created a database using the create-database.sql script as shown
in Example 6-17.

Chapter 6. Database conversion 145

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

Example 6-17 Create-database.sql script

-- Create the Initial Database
create database invent
automatic storage yes ON '/home/db2inst1/invent;
update db cfg for invent using AUTO_MAINT ON;

connect to invent;

-- Create a Bufferpool with 8K and 16K Pages

create bufferpool bp8k size 2000 pagesize 8192;
create bufferpool bp16k size 1000 pagesize 16384;

-- Make the Bufferpool Change Take Effect

disconnect invent;
connect to invent;

-- Create a Tablespace with 8K and 16K Pages

create regular tablespace tblsp8k pagesize 8192 managed by automatic storage
extentsize 8
prefetchsize automatic
bufferpool bp8k;

reate regular tablespace tblsp16k pagesize 16384 managed by automatic storage

extentsize 8
prefetchsize automatic
bufferpool bp16k;

-- Create a System Temporary Tablespace with 8K and 16K Pages

create system temporary tablespace temp8k pagesize 8 K managed by automatic
storage
extentsize 32
prefetchsize automatic
bufferpool bp8k;

create system temporary tablespace temp16k pagesize 16 K managed by automatic

storage
extentsize 32
prefetchsize automatic
bufferpool bp16k;

disconnect invent;

To create database, invoke the above SQL script from DB2 Command Line
Window or bash shell by using.
db2>@create-database.sql
bash>db2 -f create-database.sql

146 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

We then pick up database object creation statements from the output of

mysqldump, mysqlobjects.ddl file and create db2objects.ddl script. Change the
DDL statements and data types following the discussion in section 6.1 and 6.2 of
this chapter. You can create any additional statement required.

Now we execute the above created DDL scripts from bash shell as shown in
Example 6-18.

Example 6-18 Create database object form bash shell

bash>db2 CONNECT to invent
bash>db2 -tf db2objects.ddl
bash>db2 DISCONNECT invent

6.4.3 Metadata transport

In this section we discuss the usage of the database modeling tool for the
conversion of a database structure from MySQL to DB2. There are a number of
modeling tools that exist in the market that support model capturing of a MySQL
database. Most of the tools use a logical model to define the database design
and a physical model map to the target database.
򐂰 IBM Rational Rose® Professional Data Modeler Edition
The IBM Rational Rose Data Modeler Edition is a data design tool which
integrates application design with database design and maps the data model
with the object model. It allows database designers, business analysts and
developers to work together through a common language. Use this tool for
conversions that require your MySQL model file to be available for a
conversion to a DB2 physical model.

Database structure conversion process using modeling tools

Database structure conversion using the modeling tool is a very neat technique
for database conversion since many applications already have entity-relationship
diagrams. The conversion can be done very easily by following the steps given
below:
򐂰 Reverse-engineer database objects using a DDL script or an existing
database using one of the modeling tools.
򐂰 Switch to a physical model.
򐂰 Select DB2 database as a target database and generate a DDL script for the
new target.
򐂰 Create DB2 database structure using DDL scripts.

Chapter 6. Database conversion 147

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

Although this is a good technique, it is very costly since it requires modeling tools
for the conversions.

6.5 Sample database conversion

This section demonstrates the database structure conversion for our sample
Inventory Management application from MySQL to DB2. The Inventory database
is a small database consisting of tables, views, indexes, triggers, and a stored
procedure. Before starting the database conversion, see the existing database
structure in 4.2, “Database structure” on page 84.

For demonstration purposes we describe a step-by-step process using the IBM

Data Movement Tool and then describe manual method for objects, which are
not converted with the IBM Data Movement Tool and additional database
enhancements.

6.5.1 Converting database objects using the IBM Data Movement

Tool
The following steps describe the process of converting the MySQL sample
Inventory database to DB2 using the IBM Data Movement Tool:

STEP 1: Create the DB2 Database

DBAs normally like to create their database according to their storage paths
information. Therefore prior to using the IBM Data Movement Tool you need to
create the database manually. You should consider reading the IBM best
practice papers at:
http://www.ibm.com/developerworks/data/bestpractices/

For this example we use the following command to create our new DB2
database:
db2> CREATE DATABASE invent AUTOMATIC STORAGE YES ON
'/home/db2inst1/invent'

STEP 2: Specify source, connection information, and test

connection
This step is to specify source and DB2 database server connection information
and test connection within the IBM Data Movement Tool:
򐂰 Starting the IBM Data Movement Tool:

148 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

For our conversion scenario we use the IBM Data Movement Tool GUI.
<IBM Data Movement Tool Installation directory>/IBMDataMovementTool.sh
db2inst1@db2server:/opt/ibm/IBMDataMovementTool>
./IBMDataMovementTool.sh
This launches IBM Data Movement window as shown in Figure 6-8.

Figure 6-8 IBM Data Movement Tool startup window

You can also run the IBM Data Movement Tool in command line mode. The
tool will automatically switch to command line mode if it is not able to start the
GUI. If you want to run the tool in command line mode you can run the
following command:
<IBM Data Movement Tool Installation directory>/IBMDataMovementTool.sh
-console
db2inst1@db2server:/opt/ibm/IBMDataMovementTool>
./IBMDataMovementTool.sh -console
򐂰 Specifying the source and target database information:
Enter the connection information in the “Source Database” and “DB2
Database” fields. Since the IBM Data Movement Tool used in this example is
installed on the DB2 server, we connect to localhost for the DB2 server and
specifying the MySQL server IP address. You must know the following
information:
– IP Address or host name of the source and DB2 servers
– Port numbers to connect

Chapter 6. Database conversion 149

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

– Name of the databases

– A User ID with DBA privileges on the source database
– Password for that user
– Location of your source database and DB2 JDBC drivers
– Enough space or volume/mount point information where data will be
stored
򐂰 Testing connection
Once you have filled out the source and target database connection
information, the next step is to test the connection to the source and target
databases. Click Connect to MySQL to test the connection and Connect to
DB2 to test the DB2 connection. Connection results will be displayed in the
bottom left corner of the window.
Figure 6-9 on page 150 shows the values filled out for our sample conversion
and the buttons to click to test the connections.

Figure 6-9 Testing database connection using IBM Data Movement Tool

STEP 3: Extract DDL from source database

In this step of the conversion process we extract the DDL and data from the
source MySQL database. The IBM Data Movement Tool will connect to the
specified source database, extract the DDL and data, format the DDL scripts to
DB2 syntax, and generate conversion scripts.

150 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

򐂰 Specify the output directory

Specify the working directory where the DDL and data are to be extracted.
The user running the IBM Data Movement tool must have full access to the
directories specified in this step. For our sample conversion we use the
default directory:
<IBM Data Movement Tool Installation directory>/migr
򐂰 Choose whether DDL and/or DATA is required
With the IBM Data Movement Tool you have the option to extract only the
database objects, the data or both. You can do this by selecting or
unselecting the DDL and DATA Movement checkboxes. We extract only the
DDL at this point.
򐂰 Extract DDL and Data
Click the Extract Data button to extract the DDL and/or DATA and
automatically convert to DB2 syntax. Five threads are used to complete the
extractions. For data movement using the command line approach you can
control the number of treads by modifying the NUM_THREADS in unload
script. You can monitor progress in the console window, see Figure 6-10.

Chapter 6. Database conversion 151

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

Figure 6-10 DDL and data extraction output

After the data extraction has completed, go through the result output files for the
status of the data movement, warnings, errors and other potential issues. You
can click on the View Script/Output button from the Extract/Deploy window to
check the generated scripts, DDL, data or the output log file.

Table 6-3 shows the command scripts that are re-generated each time you run
the tool in GUI mode. These scripts can also be issued in console mode without
the GUI. This is helpful when you want to embed this tool as part of a batch
processes to accomplish an automated data movement.

Table 6-3 IBM Data Movement Tool scripts

Filename Description

IBMExtract.properties This file contains all input parameters that you specified
through your GUI or command line input values. You can edit
this file manually to modify or correct parameters. Note this
file will be overwritten each time you run the GUI.

152 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Filename Description

geninput This is the first data movement step where you will create an
input file that contains the names of the tables to move. You
can edit this file manually to exclude tables you do not want to
move.

genddl This is an optional script only generated if you chose to

separate DDL generation from DATA. This is the second step
in data movement to generate all DDLs from your source
database to DB2.

unload This is the last step of data movement. This script unloads
data from the source database server to flat files. DB2 LOAD
scripts will be generated after running this script. Note that if
you did not choose to separate DDL from DATA, the genddl
content is included in the unload script.

rowcount This file will be used after you have moved the data to do a
sanity check for the row count for tables in source and target
database servers.

STEP 4: Deploying database objects

Now that we have the database object definitions extracted we have three
different options to deploy the database objects to DB2:
򐂰 Deploy DDL and DATA in one step using the Deploy DDL/DATA button from
the GUI screen.
򐂰 Deploy DDL and DATA in one step by using the command line script unload.
򐂰 Deploy the database objects one by one by selecting the Interactive Deploy
tab.

Choose an option that works best for your conversion project. The interactive
deployment mode is best for deploying database objects that contain triggers,
functions and/or procedures. In most MySQL conversions the first two options
will suffice, since the conversion of the database objects will be preformed
outside of the IBM Data Movement tool. For our example, we select the
Interactive Deploy mode to better explain the conversion process by separating
the database object and data deployment. Figure 6-11 on page 154 shows the
Interactive Deployment window.

Chapter 6. Database conversion 153

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

Figure 6-11 DDL and data extraction output

From the Interactive Deploy window you can perform a number of tasks
򐂰 Refresh Database Object List
Select the refresh button (circled in Figure 6-12 on page 155) to refresh the
list of database objects in the Database Objects view on the left side of the
window.
򐂰 Edit the Object Definition
You can select the database object you would like to modify and edit in the
right side panel, as shown in Figure 6-12 on page 155. To save and deploy
changes, deploy object before selecting a new object. After deployment you
can return to refine any objects that failed to deploy.

154 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

Figure 6-12 Editing objects

You can also edit the scripts that were extracted into the conversion directory.
To change the table definition, edit the db2tables.sql file, as follows
db2inst1@db2server:/opt/ibm/IBMDataMovementTool/migr> vi db2tables.sql
Example 6-19 shows the converted DB2 table creation file.

Example 6-19 Output DB2 statements - db2tables.sql

CONNECT TO INVENT;
-- Approximate Table Size 70
--#SET :TABLE:ADMIN:GROUPS
CREATE TABLE "ADMIN"."GROUPS"
(
"GROUPNAME" VARCHAR(30) NOT NULL ,
"EDITUSER" VARCHAR(10) ,
"EDITGRANTUSERPERM" VARCHAR(10) ,
"EDITINVT" VARCHAR(10) ,
"EDITSERVICE" VARCHAR(10)
)
;

-- Approximate Table Size 48

--#SET :TABLE:ADMIN:LOCATIONS

Chapter 6. Database conversion 155

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

CREATE TABLE "ADMIN"."LOCATIONS"

(
"ID" INT NOT NULL ,
"ROOMNAME" VARCHAR(20) NOT NULL ,
"FLOORNUM" INT NOT NULL ,
"PASSCODE" VARCHAR(20)
)
;

-- Approximate Table Size 47

--#SET :TABLE:ADMIN:SEVERITY
CREATE TABLE "ADMIN"."SEVERITY"
(
"ID" INT NOT NULL ,
"TITLE" VARCHAR(15) NOT NULL ,
"NOTES" VARCHAR(20) ,
"ESTDAYS" INT ,
"AVGDAYS" INT
)
;

-- Approximate Table Size 89

--#SET :TABLE:ADMIN:INVENTORY
CREATE TABLE "ADMIN"."INVENTORY"
(
"ID" INT NOT NULL ,
"ITEMNAME" VARCHAR(15) NOT NULL ,
"MANUFACTURER" VARCHAR(20) ,
"MODEL" VARCHAR(20) ,
"YEAR" SMALLINT ,
"SERIAL" VARCHAR(20) ,
"LOCID" INT ,
"OWNERID" INT
)
;

-- Approximate Table Size 64

--#SET :TABLE:ADMIN:SERVICES
CREATE TABLE "ADMIN"."SERVICES"
(
"ID" INT NOT NULL ,
"INVENTID" INT NOT NULL ,
"DESCRIPTION" VARCHAR(20) NOT NULL ,
"SEVERITY" INT NOT NULL ,
"SERVICEOWNER" INT NOT NULL ,
"OPENDATE" DATE NOT NULL ,
"CLOSEDATE" DATE ,
"TARGETCLOSEDATE" DATE ,
"STATUS" INT
)
;

-- Approximate Table Size 39

--#SET :TABLE:ADMIN:STATUS

156 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

CREATE TABLE "ADMIN"."STATUS"

(
"ID" INT NOT NULL ,
"TITLE" VARCHAR(15) NOT NULL ,
"NOTES" VARCHAR(20)
)
;

-- Approximate Table Size 212

--#SET :TABLE:ADMIN:OWNERS
CREATE TABLE "ADMIN"."OWNERS"
(
"ID" INT NOT NULL ,
"FIRSTNAME" VARCHAR(20) NOT NULL ,
"LASTNAME" VARCHAR(20) NOT NULL ,
"EMAIL" VARCHAR(50) ,
"LOCID" INT ,
"CUBENUM" INT ,
"PHONENUM" VARCHAR(20) ,
"LOGINNAME" VARCHAR(20) NOT NULL ,
"PASSWORD" VARCHAR(20) NOT NULL ,
"FAXNUM" VARCHAR(20) ,
"GROUPS" VARCHAR(30)
)
;

TERMINATE;

Note: You should not reduce the size of any field because it may cause an
error while converting the data

򐂰 Deploy objects to target

The final step of the database conversion is to deploy the database objects.
This is performed using the Deploy All button (circled red in Figure 6-13) to
deploy all objects in the database object view or Deploy Selected button
(circled blue in Figure 6-13) to deploy the database object currently selected.
The status of the database object deployment and error messages can be
viewed in the bottom right panel, as shown in Figure 6-13.

Chapter 6. Database conversion 157

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

Figure 6-13 All database objects successfully deployed

6.5.2 Manual database object conversion and enhancements

Now that we have converted the main database objects to DB2 using the IBM
Data Movement Tool, we can use manual methods for the objects which are not
converted with the tool and add any additional database improvements.

Views
Manual conversion is required to port views from MySQL to DB2. The MySQL
view definition can be extracted from the MySQL database using the mysqldump
utility or selecting from the INFORMATION_SCHEMA.VIEWS table.

The syntax for a view in MySQL and DB2 is very similar, which makes it simple to
convert the DDL for this database object. Example 6-20 shows the CREATE
VIEW syntax for the MySQL views.

For our example, we alter the CREATE VIEW commands to match DB2 syntax,
as shown in Example 6-21.

Example 6-20 MySQL create view statement

CREATE VIEW managerGroup AS
SELECT o.id as newID, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,

158 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

o.phoneNum,
(SELECT count(*) FROM inventory i, owners o
WHERE o.id = newID and o.id = i.ownerID) as inventNum
FROM owners o, locations l
WHERE o.groups = 'manager' and o.locID = l.id;

CREATE VIEW empGroup AS

SELECT o.id as newID, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM inventory i, owners o
WHERE o.id = newID and o.id = i.ownerID) as inventNum
FROM owners o, locations l
WHERE o.groups = 'emp' and o.locID = l.id;

CREATE VIEW techGroup AS

SELECT o.id as newID, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM inventory i, owners o
WHERE o.id = newID and o.id = i.ownerID) as inventNum
FROM owners o, locations l
WHERE o.groups = 'tech' and o.locID = l.id;

CREATE VIEW bossGroup AS

SELECT o.id as newID, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM inventory i, owners o
WHERE o.id = newID and o.id = i.ownerID) as inventNum
FROM owners o, locations l
WHERE o.groups = 'boss' and o.locID = l.id;

CREATE VIEW generalGroup AS

SELECT o.id as newID, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM inventory i, owners o
WHERE o.id = newID and o.id = i.ownerID) as inventNum
FROM owners o, locations l
WHERE o.groups = 'general' and o.locID = l.id;

Example 6-21 DB2 create view statement

CREATE VIEW admin.managerGroup AS
SELECT o.id, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM admin.inventory in, admin.owners u
WHERE u.id = o.id and u.id = in.ownerID) AS inventNum
FROM admin.owners o, admin.locations l
WHERE o.groups = 'manager' and o.locID = l.id;

CREATE VIEW admin.empGroup AS

SELECT o.id, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM admin.inventory in, admin.owners u
WHERE u.id = o.id and u.id = in.ownerID) AS inventNum
FROM admin.owners o, admin.locations l
WHERE o.groups = 'emp' and o.locID = l.id;

CREATE VIEW admin.techGroup AS

SELECT o.id, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM admin.inventory in, admin.owners u
WHERE u.id = o.id and u.id = in.ownerID) AS inventNum
FROM admin.owners o, admin.locations l

Chapter 6. Database conversion 159

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

WHERE o.groups = 'tech' and o.locID = l.id;

CREATE VIEW admin.bossGroup AS

SELECT o.id, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM admin.inventory in, admin.owners u
WHERE u.id = o.id and u.id = in.ownerID) AS inventNum
FROM admin.owners o, admin.locations l
WHERE o.groups = 'boss' and o.locID = l.id;

CREATE VIEW admin.generalGroup AS

SELECT o.id, o.firstName, o.lastName, o.loginName, l.roomName, l.floorNum,
o.phoneNum,
(SELECT count(*) FROM admin.inventory in, admin.owners u
WHERE u.id = o.id and u.id = in.ownerID) AS inventNum
FROM admin.owners o, admin.locations l
WHERE o.groups = 'general' and o.locID = l.id;

Trigger conversion
Triggers also require manual conversion to port them from MySQL to DB2. The
MySQL trigger definition can be extracted from the MySQL database using the
mysqldump utility or selecting from the INFORMATION_SCHEMA.TRIGGERS
table.

Example 6-22 shows the MySQL CREATE TRIGGER statement and

Example 6-24 shows the DB2 CREATE TRIGGER statement. In these examples,
note the difference between new and old data values that are declared and
referenced. refer to 6.2.4, “Triggers manipulation” on page 137 for more details
on syntax differences between MySQL and DB2.

You should also note the change in the date function. You must find the
equivalent functions in DB2 when converting DDL and DML statements that
contain built-in functions. MySQL and DB2 built-in functions and operators are
discussed and compared in more detail in 8.1.10, “Built-in functions and
operators” on page 223.

We have found a similar function in DB2 to replace the curDate function in

MySQL. However, the default date format needs to be modified to ISO
(YYYY-MM-DD). This is because the default format for the DB2 current date
function is MM/DD/YY, whereas the MySQL curDate function is YYYY-MM-DD.
The commands described in Example 6-23 on page 161 change the default
format to ISO (YYYY-MM-DD).

Example 6-22 MySQL updateDate Trigger

CREATE TRIGGER updateDate BEFORE UPDATE on services
FOR EACH ROW
BEGIN
IF NEW.status = 7 THEN

160 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

SET NEW.closeDate = CURDATE();

END IF;
END

Example 6-23 Commands to change to ISO format

db2inst1@db2server:~> cd /home/db2inst1/sqllib/bnd/
db2inst1@db2server:~/sqllib/bnd> db2 CONNECT TO invent

Database Connection Information

Database server = DB2/LINUX 9.7.0

SQL authorization ID = DB2INST1
Local database alias = INVENT

db2inst1@db2server:~/sqllib/bnd> db2 bind @db2ubind.lst datetime ISO blocking

all grant public

Example 6-24 DB2 updateDate trigger

CREATE TRIGGER updateDate BEFORE UPDATE on admin.services
REFERENCING NEW AS N OLD AS O
FOR EACH ROW
BEGIN
IF N.status = 7 THEN
SET N.closeDate = CURRENT DATE;
END IF;
END

Stored procedure conversion

Manual conversion is required to port stored procedures from MySQL to DB2.
The MySQL stored procedure definition can be extracted from the MySQL
database by selecting the definition from the
INFORMATION_SCHEMA.ROUTINES table.

The syntax for a procedure in MySQL and DB2 is similar. Example 6-25 on
page 161 shows the CREATE PROCEDURE syntax for the MySQL procedure
and Example 6-26 on page 162 shows the CREATE PROCEDURE for the DB2
procedure. You may notice the only difference between the statements the date
function to determine the number of days between the open date and the close
date. For a description of MySQL built-in functions and DB2 equivalent functions
or solution refer to 8.1, “Data Manipulation Language differences and similarities”
on page 208.

Example 6-25 MySQL create updateAvgDays procedure

CREATE PROCEDURE updateAvgDays(IN sevLevel INT)

Chapter 6. Database conversion 161

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

BEGIN
UPDATE severity SET avgDays = (SELECT SUM(Datediff(closeDate,
openDate))/Count(*)
FROM services where severity = sevLevel and status = 7)
WHERE id = sevLevel;
END

Example 6-26 DB2 create updateAvgDays procedure

CREATE PROCEDURE updateAvgDays(IN sevLevel INT)
BEGIN
UPDATE admin.severity
SET avgDays = (SELECT SUM(TIMESTAMPDIFF(16,CHAR(TIMESTAMP(closeDate) -
TIMESTAMP(openDate))))/count(*)
FROM admin.services WHERE severity = sevLevel and status = 7)
WHERE id = sevLevel;
END

Foreign keys
Now any additional enhancement can be added to your database using the DB2
features that may have not been supported in your existing MySQL storage
engine. One example is referential integrity. This is essential to the database by
ensuring consistency of data values between related columns in different tables.
Referential integrity is usually maintained by using the primary key, unique key,
and foreign keys. MySQL only supports foreign keys in the InnoDB engine.
Primary and unique are successfully converted using the IBM Data Movement
Tool but at the time of writing this book foreign keys are not supported for a
MySQL conversion project. If you want create foreign keys in your database or
convert your foreign keys from an Innodb database this needs to add manually.

In our sample application, we created referential integrity between tables in a file

the referential.ddl file, which looks like Example 6-27.

Example 6-27 Add foreign keys

ALTER TABLE admin.owners ADD CONSTRAINT OWNERLOCI FOREIGN KEY (locid)
REFERENCES admin.locations (id) ON DELETE CASCADE;

ALTER TABLE admin.owners ADD CONSTRAINT OWNERGRPI FOREIGN KEY (groups)

REFERENCES admin.groups (GROUPNAME) ON DELETE CASCADE;

ALTER TABLE admin.inventory ADD CONSTRAINT INVENTLOCI FOREIGN KEY (locid)

REFERENCES admin.locations (id) ON DELETE CASCADE;

ALTER TABLE admin.inventory ADD CONSTRAINT INVENTOWNERI FOREIGN KEY (ownerid)

REFERENCES admin.owners (id) ON DELETE CASCADE;

162 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

ALTER TABLE admin.services ADD CONSTRAINT SERVINVENTI FOREIGN KEY (inventid)

REFERENCES admin.inventory (id) ON DELETE CASCADE;

ALTER TABLE admin.services ADD CONSTRAINT SERVOWNERI FOREIGN KEY (serviceowner)

REFERENCES admin.owners (id) ON DELETE CASCADE;

ALTER TABLE admin.services ADD CONSTRAINT SERVSERVERITYI FOREIGN KEY (severity)

REFERENCES admin.severity (id) ON DELETE CASCADE;

ALTER TABLE admin.services ADD CONSTRAINT SERVSTATI FOREIGN KEY (status)

REFERENCES admin.status (id) ON DELETE CASCADE;

Now we have completed the DDL modification; we executed the above changed
scripts to create the DB2 database and the objects as shown in Example 6-28.

Example 6-28 creation of tables and databases in DB2

db2inst1@db2server:~/DB2Scripts> db2 CONNECT TO invent

Database Connection Information

Database server = DB2/LINUX 9.5.0

SQL authorization ID = DB2INST1
Local database alias = INVENT

db2inst1@db2server:~/DB2Scripts> db2 -tf referential.ddl

DB20000I The SQL command completed successfully.
DB20000I The SQL command completed successfully.
DB20000I The SQL command completed successfully.
DB20000I The SQL command completed successfully.
DB20000I The SQL command completed successfully.
DB20000I The SQL command completed successfully.
DB20000I The SQL command completed successfully.
DB20000I The SQL command completed successfully.

Automatic maintenance
Performing maintenance activities on your databases is essential to ensuring
that they are optimized for performance and recoverability. The database
manager provides automatic maintenance capabilities for performing database
backups, keeping statistics current and reorganizing tables and indexes as
necessary.

For users, it can be time-consuming to determine when to run maintenance

activities. Automatic maintenance takes the burden off users. With automatic
maintenance, you can specify your maintenance objectives, including when
automatic maintenance can run. The DB2 database manager uses the objectives
you have specified to determine if the maintenance activities need to be done

Chapter 6. Database conversion 163

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

and runs only the required maintenance activities during the next available
maintenance window (a user-defined time period for the running of automatic
maintenance activities). Example 6-29 shows the activities that can be controlled
by the database manager’s automatic maintenance feature.

Example 6-29 Automatic maintenance database manager variables

Automatic maintenance (AUTO_MAINT) = ON
Automatic database backup (AUTO_DB_BACKUP) = OFF
Automatic table maintenance (AUTO_TBL_MAINT) = ON
Automatic runstats (AUTO_RUNSTATS) = ON
Automatic statement statistics (AUTO_STMT_STATS) = OFF
Automatic statistics profiling (AUTO_STATS_PROF) = OFF
Automatic profile updates (AUTO_PROF_UPD) = OFF
Automatic reorganization (AUTO_REORG) = ON

Enablement of the automatic maintenance features is controlled using the

automatic maintenance database configuration parameters. These are a
hierarchical set of switches to allow for simplicity and flexibility in managing the
enablement of these features. You can automate database maintenance
activities to run only when they are needed using the Configure Automatic
Maintenance wizard. The DB2 database manager uses the objectives you have
specified using the Configure Automatic Maintenance wizard to determine
whether the maintenance activities need to be done. Then the DB2 database
manager runs only the required maintenance activities during the next available
maintenance window.

Self-Tuning Memory Manager

A revolutionary memory tuning system called the Self-Tuning Memory Manager
(STMM) was introduced in DB2 9. It works on main database memory
parameters including buffer pools, sort heaps, locklist, package cache, and total
database memory. It allows hands-off online memory tuning without DBA
intervention by sensing the underlying workload and tunes the memory based on
needs. When workload shifts, and memory redistribution is required to achieve
optimal performance, STMM can adapt quickly to adjust memory configuration.

Self tuning memory simplifies the task of memory configuration by automatically

setting values for memory configuration parameters and sizing buffer pools.
When enabled, the memory tuner dynamically distributes available memory
resources between several memory consumers including sort, package cache
and lock list areas and buffer pools. Example 6-30 shows how to enable STMM
and enable specific parameters be automatically tuned.

Example 6-30 Setting up Self-Tuning Memory Manager

db2> UPDATE DATABASE CONFIGURATION FOR invent SELF_TUNING_MEM ON

164 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch06.fm

db2> UPDATE DB CFG FOR invent USING locklist AUTOMATIC

maxlocks AUTOMATIC
pckcachesz AUTOMATIC
sheapthres_shr AUTOMATIC
sortheap AUTOMATIC

Chapter 6. Database conversion 165

7093ch06.fm Draft Document for Review October 16, 2009 10:22 am

166 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

Chapter 7. Data conversion

There are various considerations around converting data from a MySQL
database to DB2. This chapter we focus on several of these and describe the
usage of tools and commands, aiding in data movement from a MySQL database
to a DB2 database.

This chapter also discusses the differences in specific data formats and data
types and ways in which they can be converted from MySQL to DB2.

We also describe how user account management (user data, access rights, and
privileges) is implemented in MySQL and how this information can be ported to
implement secure database access within DB2.

Finally, the steps for how we did the data conversion in our sample project are
discussed in detail.

© Copyright IBM Corp. 2009. All rights reserved. 167

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

7.1 Data porting considerations

Data porting describes the necessary steps to get data from one database to
another. In general, you have to unload (also referred to as dump or export) the
data from the source database into one or more files and load (also referred to as
import) these files into the target database.

Database systems provide commands and tools for unloading and loading data.
In MySQL the mysqldump tool is used to unload a database. DB2 provides the
LOAD and IMPORT commands for loading data from files into the database.

You have to be aware of the differences in how specific data types are
represented by different database systems. For example, the representation of
date and time values may be different in different databases, and this often
depends on the local settings of the system.

If the source and the target databases use different formats, you must to convert
the data either automatically using tools or manually. Otherwise, the loading tool
would not be able to understand the data it needs to load due to improper
formatting.

The conversion of binary data stored in BLOBs should be done manually

because binary data cannot be exported to files in a text format.

Porting the user account management is a very specific step in a conversion

project. You must get the information about users, access rights, and privileges
out of MySQL, convert it to DB2 specific security information, and create the
users in DB2 according to the source data. Porting encrypted passwords is
impossible in most cases.

7.1.1 Data porting commands and tools

MySQL and DB2 both provide tools that support data porting between the two
systems. In MySQL the mysqldump utility and the mysqlhotcopy scripts are
used to retrieve data from the database; the LOAD and IMPORT commands can be
used to get this data into DB2.

The IBM Data Movement Tools automates the use of the MySQL SELECT
statements and the DB2 LOAD commands.

The mysqldump tool

When porting data, this tool can be used to retrieve the data from MySQL tables.
It is shipped with MySQL and is usually located in the bin directory of the MySQL
installation.

168 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

The basic usage of the mysqldump tool is:

mysqldump [OPTIONS] database [tables]

For a complete description of this tool, run mysqldump --help. Some important
command line options to keep in mind are:
򐂰 --no-data
This option ensures that no data is extracted from the database, just the SQL
statements for creating the tables and indexes, therefore, used for extracting
DDL statements only.
򐂰 --no-create-info
This option ensures that no SQL statements for creating the exported table
are extracted from the database. Therefore, it is used for exporting data only.
The output file containing the data can be loaded into a DB2 table at a later
time.
򐂰 --tab=<outFilePath>
This option creates a text file with the DDL (<tablename>.sql) and a tab
separated text file with the data (<tablename>.txt) in the given path for each
specified table. This option works only when the utility is run on the same
machine as the MySQL daemon. If this option is not specified, INSERT
statements for each row are created.

Example 7-1 shows the usage and output of the mysqldump command using only
the --user and --password options. The output includes DDL statements for table
creation and INSERT statements to insert data into the table.

Example 7-1 Usage of mysqldump with only the --user and --password options
mysqlServer:~ # mysqldump --user root --password inventory severity

-- MySQL dump 10.13 Distrib 5.1.36, for pc-linux-gnu (i686)

--
-- Host: localhost Database: inventory
-- ------------------------------------------------------
-- Server version 5.1.36-log

--
-- Table structure for table `severity`
--

DROP TABLE IF EXISTS `severity`;

CREATE TABLE `severity` (

`id` int(15) NOT NULL,
`title` varchar(15) NOT NULL,
`notes` varchar(20) DEFAULT NULL,

Chapter 7. Data conversion 169

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

`estDays` int(20) DEFAULT '14',

`avgDays` int(20) DEFAULT NULL,
PRIMARY KEY (`id`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

--
-- Dumping data for table `severity`
--

LOCK TABLES `severity` WRITE;

INSERT INTO `severity` VALUES

2,'high-med',NULL,4,4),
(4,'low-med',NULL,10,10),
(3,'medium',NULL,7,7),(
5,'low',NULL,14,12),(
1,'high',NULL,1,2);

UNLOCK TABLES;

-- Dump completed on 2009-08-25 2:11:48

Example 7-2 shows the usage and output of the mysqldump command with the
--no-create-info but without the --tab options. The output has only INSERT
statements for each row to insert data into the table.

Example 7-2 Usage of mysqldunp with the --no-create-info but without the --tab option
mysqlServer:~# mysqldump--no-create-info --user root -password inventory severity

-- MySQL dump 10.13 Distrib 5.1.36, for pc-linux-gnu (i686)

--
-- Host: localhost Database: inventory
-- ------------------------------------------------------
-- Server version 5.1.36-log

--
-- Dumping data for table `severity`
--

LOCK TABLES `severity` WRITE;

INSERT INTO `severity` VALUES

(2,'high-med',NULL,4,4),
(4,'low-med',NULL,10,10),
(3,'medium',NULL,7,7),
(5,'low',NULL,14,12),
(1,'high',NULL,1,2);

UNLOCK TABLES;

170 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

-- Dump completed on 2009-08-25 2:28:23

Example 7-3 shows the usage and output of the mysqldump command with the
--no-create-info and the --tab options. This outputs a file in the current directory
named <tableName>.txt that contains only the exported MySQL data. This file
can be read by the DB2 LOAD command.

Example 7-3 Usage of mysqldump with the --no-create-info and the --tab option
mysqlServer:~ # mysqldump --no-create-info --tab=. --user root --password
inventory severity

mysqlServer:~ # cat severity.txt

2 high-med \N 4 4
4 low-med \N 10 10
3 medium \N 7 7
5 low \N 14 12
1 high \N 1 2

Example 7-4 shows the usage and output of mysqldump without the
--no-create-info but with the --tab options. The output are two files: one
containing the DDL statements for table creation (<tableName>.sql), the other
one with the exported MySQL data (<tableName>.txt) in the current directory.
The second file can be read by the DB2 LOAD command.

Example 7-4 Usage of mysqldump without the --no-create-info but with the --tab option
mysqlServer:~ # mysqldump --tab=. --user root -password inventory severity

mysqlServer:~ # cat severity.sql

-- MySQL dump 10.13 Distrib 5.1.36, for pc-linux-gnu (i686)

--
-- Host: localhost Database: inventory
-- ------------------------------------------------------
-- Server version 5.1.36-log

--
-- Table structure for table `severity`
--

DROP TABLE IF EXISTS `severity`;

CREATE TABLE `severity` (

`id` int(15) NOT NULL,
`title` varchar(15) NOT NULL,
`notes` varchar(20) DEFAULT NULL,
`estDays` int(20) DEFAULT '14',
`avgDays` int(20) DEFAULT NULL,

Chapter 7. Data conversion 171

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

PRIMARY KEY (`id`)

) ENGINE=MyISAM DEFAULT CHARSET=latin1;

-- Dump completed on 2009-08-25 2:49:30

mysqlServer:~ # cat severity.txt

2 high-med \N 4 4
4 low-med \N 10 10
3 medium \N 7 7
5 low \N 14 12
1 high \N 1 2

The mysqlhotcopy script

When porting data, the mysqlhotcopy Perl script can be used to retrieve the data
from MySQL tables. It is shipped with MySQL, located in the bin directory of the
MySQL installation and requires the Perl DBI module installed.

The basic usage of the mysqlhotcopy script is as follows:

mysqlhotcopy database [/path/]

IBM Data Movement Tool SELECT script

The IBM Data Movement Tool generates a file with the SELECT statement
required to extract the data from the MySQL database. This file called
<database>.tables is located in the conversion output directory and is generated
during the DDL extraction phase, described in 6.5.1, “Converting database
objects using the IBM Data Movement Tool” on page 148.

Example 7-5 illustrates the inventory.tables generated file for our sample
conversion.

Example 7-5 inventory.tables file

db2server:/opt/ibm/IBMDataMovementTool/migr # cat inventory.tables
"ADMIN"."groups":SELECT * FROM groups
"ADMIN"."inventory":SELECT * FROM inventory
"ADMIN"."locations":SELECT * FROM locations
"ADMIN"."owners":SELECT * FROM owners
"ADMIN"."services":SELECT * FROM services
"ADMIN"."severity":SELECT * FROM severity
"ADMIN"."status":SELECT * FROM status

There are a few other scripts that are generated during the DDL extraction phase
that are used to extract the data from the MySQL database and load it into DB2.
The unload script is used to unload the data from the MySQL database and store
it in <schema>_<tableName>.txt files in the <migration output directory>/data.

172 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

The db2load.sh script can then be used to load these files into the DB2
database. The db2load.sql script is executed from the db2gen.sh script which
also executes other .sql generated scripts to port the database. You can also
execute these using the GUI, which we discuss further in the chapter.

DB2 loading tools

DB2 provides two utilities for loading data into a database: LOAD and IMPORT.

In general, the LOAD utility is faster than the IMPORT utility, because it writes
formatted pages directly into the database, while the IMPORT utility performs SQL
insert statements. The LOAD utility validates the uniqueness of the indexes, but
does not fire triggers, and does not perform referential or table constraints
checking.

DB2 LOAD command

The LOAD utility is capable of efficiently moving large quantities of data into newly
created tables, or into tables that already contain data.

The load process goes through four phases:

򐂰 Load data.
򐂰 Build indexes.
򐂰 Delete rows with a unique key violation or a data link violation.
򐂰 Copy index data from the system temporary table space to the original table
space.

See Example 7-6 for a simplified syntax diagram for the LOAD command. For a
complete syntax description, visit the Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

Example 7-6 shows a simplified syntax of the DB2 LOAD command.

Example 7-6 Simplified syntax of the DB2 LOAD command

>>-LOAD--+--------+--FROM----+-filename---+-+--OF--filetype----->
'-CLIENT-' +-pipename---+
+-device-----+
'-cursorname-'
>--+-------------------------------+---------------------------->
| .--------------. |
| V | |
'-MODIFIED BY----filetype-mod-+-'
>--+--------------+--+-------------+--+-----------------+------->
'-SAVECOUNT--n-' '-ROWCOUNT--n-' '-WARNINGCOUNT--n-'
>--+------------------------+----------------------------------->
'-MESSAGES--message-file-'
>--+-------------------------------+--+-INSERT----+------------->

Chapter 7. Data conversion 173

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

'-TEMPFILES PATH--temp-pathname-' +-REPLACE---+

+-RESTART---+
'-TERMINATE-'
>--INTO--table-name--+-------------------------+---------------->
| .-,-------------. |
| V | |
'-(----insert-column-+--)-'
.-ALLOW NO ACCESS-----------------------------.
>--+---------------------------------------------+-------------->
'-ALLOW READ ACCESS--+----------------------+-'
'-USE--tablespace-name-'
>--+--------------------------------------+--------------------->
'-CHECK PENDING CASCADE--+-IMMEDIATE-+-'
'-DEFERRED--'
>--+-----------------+------------------------------------------>
'-LOCK WITH FORCE-'

The IBM Data Movement Tool uses the LOAD utility for loading the application
data into DB2. Example 7-7 shows an example of the LOAD command used to
load the data into the Severity table.

Example 7-7 IBM Data Movement Tool LOAD example

--#SET :LOAD:ADMIN:SEVERITY
LOAD FROM "/opt/ibm/IBMDataMovementTool/migr/data/admin_severity.txt"
OF DEL
MODIFIED BY CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL""
DELPRIORITYCHAR NOROWWARNINGS
--DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_severity.txt"
METHOD P (1,2,3,4,5)
MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_severity.txt"
REPLACE INTO "ADMIN"."SEVERITY"
(
"ID",
"TITLE",
"NOTES",
"ESTDAYS",
"AVGDAYS"
)
--STATISTICS YES WITH DISTRIBUTION AND DETAILED INDEXES ALL
NONRECOVERABLE
INDEXING MODE AUTOSELECT
;

DB2 IMPORT command

The IMPORT utility inserts data from an input file into a table or updateable view. If
the table or view receiving the imported data already contains data, you can
either replace or append this new data to the existing data.

174 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

See Example 7-8 for a simplified syntax diagram for the IMPORT command. For a
complete syntax description, visit the Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

Example 7-8 Simplifies syntax of the DB2 IMPORT command

>>-IMPORT FROM --filename--OF--filetype----------------------------->
>--+--------------------------------+------------------------------->
| .--------------. |
| V | |
'-MODIFIED BY----filetype-mod--+-'
>--+------------------------+--------------------------------------->
'-MESSAGES--message-file-'
>--+-+-INSERT---------+------INTO--+-table-name--+---------------------------->
|-INSERT_UPDATE----+ | | .-,------------. | |
|-+-REPLACE--------+ | | V | | |
|-+-REPLACE_CREATE-+ | '--(--insert-column-+--)-' |
| '-| hierarchy description |--------------'

7.1.2 Differences in data formats

When porting data from MySQL to DB2, specific data types require some
attention. The DATETIME and TIMESTAMP data types have the same content
in MySQL and DB2, but have a different representation. YEAR does not have a
representation in DB2 but it can be mapped as a CHAR or SMALLINT. When
loading the exported data into DB2 you have to be aware of this data format.

Binary Large Objects (BLOB data type) usually contain binary data. Exporting of
binary data into text files is not possible. So, if your BLOBs contain binary data,
you must convert them in a different way than exporting and loading. The IBM
Data Movement tool handles the conversion BLOB data for you.

MySQL DATETIME and TIMESTAMP data type format

Both MySQL data types DATETIME and TIMESTAMP must be mapped to the
DB2 data type TIMESTAMP and YEAR must be mapped to either CHAR or
SMALLINT. Therefore, exported MySQL application data of these types must be
in a DB2 readable format before it is loaded into DB2.

The mysqldump format of MySQL DATETIME and TIMESTAMP values is

“YYYY-MM-DD hh:mm:ss”, for example, “2009-08-30 14:21:14”. Notice the
separators.

The DB2 LOAD command let you specify the file-type-modifier clause
TIMESTAMPFORMAT, which determines the formatting of the TIMESTAMP
values. If you want to import MySQL TIMESTAMP values, you must change the
LOAD command in the deploy.sh script to the following:

Chapter 7. Data conversion 175

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

db2 LOAD from ""<outFilePath>/<tablename>.dat"" |

of DEL
modified by
coldel0x09
timestampformat=\" YYYY-MM-DD HH:MM:SS\"
insert into <schemaname>.<tablename>

If you are using the IBM Data Movement Tool you do not have to worry about this
format since the tool takes care of this for you.

Manipulating data extraction

In some cases you may want to manipulate the data when it is extracted from the
MySQL database. For example in our sample conversion, the application stores
the application user passwords encrypted in the owners table. Since we know
the key used to encrypt each password, these can be decrypted during
extraction. The IBM Data Movement Tool uses SELECT statements to extract
the data from the MySQL database. You can find the SELECT commands in the
<database>.tables file. To modify the extraction commands, open the file in an
editor and make the necessary changes to the SELECT command. The
highlighted text in Example 7-9 illustrates the changes we made to the
inventory.tables file.
Example 7-9 Edits made to IBM Data Movement Tool extraction script
"ADMIN"."groups":SELECT * FROM groups
"ADMIN"."inventory":SELECT * FROM inventory
"ADMIN"."locations":SELECT * FROM locations
"ADMIN"."owners":SELECT id, firstName, lastName, email, locID, cubeNum,
phoneNum, loginName, AES_DECRYPT(password,'password') as password, faxNum,
groups FROM owners
"ADMIN"."services":SELECT * FROM services
"ADMIN"."severity":SELECT * FROM severity
"ADMIN"."status":SELECT * FROM status

When extracting the data from the GUI tool make sure to select No when
requested to recreate the conversion directory and the <database>.tables file.
Otherwise your changes will be overwritten.

7.1.3 Differences in the user account management

The method in which the access rights and privileges are stored in MySQL is
completely different from DB2.

176 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

MySQL user account management

Porting the user account management from MySQL to the DB2 security system
requires knowledge about how the user account management data affects your
application, and how user data, passwords, access rights, and privileges are
stored in MySQL.

User accounts
When assessing your application, be sure to distinguish between the following
user types:
򐂰 Application user accounts
These users log onto the application, but do not exist on the database level.
Database access is through the application with the application’s database
user ID. As the information about application users is usually stored in custom
application tables, the porting of application user account data is done when
porting the MySQL application data. In the sample inventory database this
would be our owners table.
򐂰 Database user accounts
Database users connect directly to the MySQL database to retrieve and
manipulate data. At least one database user must exist for applications to
connect to the database. Database user accounts are created with the
MySQL server and allow you to grant and restrict access to portions of the
MySQL servers. A database user account is associated with a host name and
user name. The user account information is stored with the mysql.user table
and must be ported in data conversion step. Access rights and privileges for
these users are stored in the mysql.db, mysql.host, mysql.tables_priv,
mysql.columns_priv, and mysql.procs_priv tables.

Passwords
Database users have associated passwords, which are stored encrypted in the
mysql.user table. Encrypted passwords cannot be ported and must be reset on
the new system. The password of the database user, which is used by an
application to access the database, is typically stored in a profile with restricted
rights.

Note: The porting of encrypted database user passwords is impossible,

because it is the intended purpose of encryption functions to make passwords
not able to be decrypted!

Access rights
When accessing a MySQL database there are two levels of access control.
When you first connect to a MySQL server you provide a user name and the

Chapter 7. Data conversion 177

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

associated password. Furthermore, the machine connecting from must be

associated with this user to allow the connection. This is based on the
assumption that a user with a specific user name from one host is different from
a user with the same user name from a different host.

This access information is stored in the mysql.user table in the fields: user,
password, and host. The MySQL wildcard % is often used in the host field to
specify that this user can connect from any host. The wildcard underscore “_” is
also sometimes used for single characters.

In Example 7-10 the user user1 can connect from any host, the user user2 can
connect from just the host remoteHost.ibm.com, and db2inst1 can connect from
localhost, myServer and 127.0.0.1.

Example 7-10 Sample user data for connection verification

mysql> select user, password, host from mysql.user;
+---------+-------------------------------------------+--------------------+
| user | password | host |
+---------+-------------------------------------------+--------------------+
| db2inst1| EE2FD1618DEE42FD1618BB9JD33HD736H5HNNT757 | myServer |
| db2inst1| GGFD1618BDJ473SKCNGU8JB9JD33HD73J5748SUFL | 127.0.0.1 |
| user1 | 2FD162470C0C06DEE418BDCA2EC9D1E1B99005A9K | %.com |
| user2 | *2470C0C06DEE42FD1618BB99005ADCA2EC9D1E19 | remoteHost.ibm.com |
| db2inst1| 06DE005ADCA2EC2470C0C99005*ADCA2EC8BB9900 | localhost |
+---------+-------------------------------------------+--------------------+

Refer to the MySQL Reference Manual for a complete description of the MySQL
connection verification at:
http://dev.mysql.com/doc/refman/5.1/en/index.html

You can also find information on how the entries in the mysql.user table are
ordered when the provided connection data meets more than one connection
criteria.

Privileges
The second level of access control occurs once a connection to the MySQL
database is established, each time a command is run against the database,
MySQL checks if the connected user has sufficient privileges to run this
command.

Privileges exist for selecting, inserting, updating, and deleting data for creating,
altering, and dropping database objects and other operations performed at the
database level.

All privileges are stored in either the mysql.user, mysql.db, mysql.host,

mysql.tables_priv, mysql.columns_priv, or mysql.procs_priv tables.

178 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

Privileges in MySQL can be granted on the following levels:

򐂰 Global level
Global privileges provides control over the overall MySQL server and are
stored in the mysql.user table.
򐂰 Database level
Database privileges provides control over a specific database and are stored
in the mysql.db (database privileges) and mysql.host (host privileges) tables.
򐂰 Table level
Table privileges provides control over a specific table and are stored in the
mysql.tables_priv table.
򐂰 Column level
Column privileges provides control over a specific column in a table and are
stored in the mysql.columns_priv table.
򐂰 Routine level
Stored routine privileges provides control over a specific stored routine and
are stored in the mysql.procs_priv table.

Privileges can be granted to users with the MySQL GRANT command; they can be
revoked with the REVOKE command.

For more information about MySQL privileges, see the MySQL Reference
Manual, at:
http://dev.mysql.com/doc/refman/5.1/en/

DB2 security system

There are four main mechanisms within DB2 that allow DB2 to implement a
security plan: authentication, administrative authorization, privileges and
Label-Based Access Control (LBAC). DB2 works closely with the security
features of the underlying operating system to authenticate users’ specific
database privileges and authorizations are granted to operating system users,
groups, or DB2 roles.

User accounts
To create a user for DB2 implies creation of a user in the server’s operating
system, assigning the user to a group, and granting specific database privileges
to the user or group.

On Linux systems you must have root access to the system to create groups and
users. Group information is stored in the file /etc/group, user information in the
file /etc/passwd.

Chapter 7. Data conversion 179

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

For example, if you want to create a new group db2app1 with one user db2usr1
to access a specific DB2 table, the necessary steps are:
1. Log on to the Linux system with root privileges.
2. Create the group. Make sure that the provided group name does not already
exist and ensure that it is no longer than eight characters:
groupadd [-g 995] db2app1
3. Create the user and assign it to the previously created group. Make sure that
the ID for the user does not already exist and is not longer than eight
characters:
useradd [-u 1001] -g db2app1 -m -d /home/db2usr1 db2usr1 [-p passwd1]
If the user is going to access the DB2 database locally, then continue with the
next two steps:
4. Edit the profile of the created user:
vi /home/db2usr1/.profile
5. Add the following line to the profile. Be sure to specify the path of your
DB2instance owner’s home directory and to specify a blank between the dot
and the command:
. /home/db2inst1/sqllib/db2profile

DB2 9.5 introduced roles to simplify management of authorization. Roles are

equivalent of in-database groups and allow DBAs to group together one or more
privileges, authorities, or security labels. Roles can be assigned to users, groups,
PUBLIC or to other roles using the GRANT statement.

Passwords
The passwords that are used for DB2 are the system passwords of the user. To
set a password in Linux use the passwd <username> command as root user.

Access rights
The first component in the DB2 security model is authentication. Access to DB2
databases is restricted to users that exist on the DB2 system. When connecting
to a DB2 database you have to provide a user name and password that is valid
against the server's system. Authentication can occur at the DB2 server or DB2
client using operating system authentication, Kerberos or an external security
manager.

Authorities and privileges

Privileges enable users to create or access database objects. Authority levels
provide a method of grouping privileges and control over higher-level database
manager maintenance and utility operations. Label-based access control (LBAC)
provides a more granular approach to granting privileges on a row and column

180 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

basis. Together, these control access to the database manager and its database
objects. Users can access only those objects for which they have the appropriate
authorization, that is, the required privilege or authority.

Administrative authority
A user or group can have one or more of the following administrative authorities:

System-level authorization
The system-level authorities provide varying degrees of control over
instance-level functions:
򐂰 SYSADM authority
The SYSADM (system administrator) authority provides control over all the
resources created and maintained by the database manager. The system
administrator possesses all authorities for the SYSCTRL, SYSMAINT, and
SYSMON authority. The user who has SYSADM authority is responsible for
both controlling the database manager and ensuring the safety and integrity
of the data.
򐂰 SYSCTRL authority
The SYSCTRL authority provides control over operations that affect system
resources. For example, a user with SYSCTRL authority can create, update,
start, stop, or drop a database. This user can also start or stop an instance,
but cannot access table data. Users with SYSCTRL authority also have
SYSMON authority.
򐂰 SYSMAINT authority
The SYSMAINT authority controls maintenance operations on all databases
associated with an instance. A user with SYSMAINT authority can update the
database configuration, backup a database or table space, restore an existing
database, and monitor a database. Like SYSCTRL, SYSMAINT does not
provide access to table data. Users with SYSMAINT authority also have
SYSMON authority.
򐂰 SYSMON authority
The SYSMON (system monitor) authority controls the usage of the database
system monitor.

Figure 7-1 illustrates the instance level authorities that can be granted to a user
or role.

Chapter 7. Data conversion 181

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

Figure 7-1 Instance level authorities

Database-level authorization
The database level authorities provide control within a database:
򐂰 DBADM (database administrator)
The DBADM authority level provides administrative authority over a single
database. This database administrator possesses the privileges required to
create objects and issue database commands.
The DBADM authority can only be granted by a user with SECADM authority
and cannot be granted to PUBLIC.
򐂰 SECADM (security administrator)

182 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

The SECADM authority level provides administrative authority for security

over a single database. The security administrator authority possesses the
ability to manage database security objects (database roles, audit policies,
trusted contexts, security label components, and security labels) and grant
and revoke all database privileges and authorities. A user with SECADM
authority can transfer ownership of objects that they do not own. They can
also use the AUDIT statement to associate an audit policy with a particular
database or database object at the server.
The SECADM authority has no inherent privileges to access data stored in
tables. This authority can only be granted by a user with SECADM authority.
The SECADM authority cannot be granted to PUBLIC.
򐂰 SQLADM (SQL administrator)
The SQLADM authority level provides administrative authority to monitor and
tune SQL statements within a single database. It can be granted by a user
with ACCESSCTRL or SECADM authority.
򐂰 WLMADM (workload management administrator)
The WLMADM authority provides administrative authority to manage
workload management objects, such as service classes, work action sets,
work class sets, and workloads. It can be granted by a user with
ACCESSCTRL or SECADM authority.
򐂰 EXPLAIN (explain authority)
The EXPLAIN authority level provides administrative authority to explain
query plans without gaining access to data. It can only be granted by a user
with ACCESSCTRL or SECADM authority.
򐂰 ACCESSCTRL (access control authority)
The ACCESSCTRL authority level provides administrative authority to issue
the following GRANT (and REVOKE) statements. ACCESSCTRL authority can
only be granted by a user with SECADM authority. The ACCESSCTRL
authority cannot be granted to PUBLIC.
– GRANT (Database authorities)
– GRANT (Global variable privileges)
– GRANT (Index privileges)
– GRANT (Module privileges)
– GRANT (Package privileges)
– GRANT (Routine privileges)
– GRANT (Schema privileges)
– GRANT (Sequence privileges)
– GRANT (Server Privileges)
– GRANT (Table, view, or nickname privileges)
– GRANT (Table space privileges)
– GRANT (Workload privileges)

Chapter 7. Data conversion 183

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

– GRANT (XSR object privileges)

򐂰 DATAACCESS (data access authority)
The DATAACCESS authority level provides the following privileges and
authorities. It can be granted only by a user who holds SECADM authority.
The DATAACCESS authority cannot be granted to PUBLIC.
– LOAD authority
– SELECT, INSERT, UPDATE, DELETE privilege on tables, views,
nicknames, and materialized query tables
– EXECUTE privilege on packages
– EXECUTE privilege on modules
– EXECUTE privilege on routines
򐂰 Database authorities (non-administrative)
To perform activities such as creating a table or a routine, or for loading data
into a table, specific database authorities are required. For example, the
LOAD database authority is required for use of the load utility to load data into
tables (a user must also have INSERT privilege on the table).

Figure 7-2 on page 185 illustrates the database level authorities that can be
granted to a user or role.

184 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

Figure 7-2 Database level authorities

Chapter 7. Data conversion 185

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

Privileges
A privilege is a permission to perform an action or task. Authorized users can
create objects, have access to objects they own, and can pass on privileges on
their own objects to other users by using the GRANT statement.

Privileges may be granted to individual users, to groups, or to PUBLIC. PUBLIC

is a special group that consists of all users, including future users. Users that are
members of a group will indirectly take advantage of privileges granted to the
group, where groups are supported.

Privileges can be assigned to a user in the following three ways:

򐂰 Explicit
Individual privileges can be explicitly granted to allow a user or group to carry
out specific tasks on specific objects. These privileges can be given and
revoked using the GRANT and REVOKE commands. Here are some examples:
db2 GRANT SELECT on TABLE db2inst1.t1 to employee
db2 REVOKE SELECT on TABLE db2inst1.t1 from employee
Users with the administrative authorities ACCESSCTRL or SECADM, or with
the CONTROL privilege, can grant and revoke privileges to and from users.
Individual privileges and database authorities allow for a set of activities, but
do not include the right to grant the same privileges or authorities to other
users. The right to grant table, view, schema, package, routine, and
sequence privileges to others can be extended to other users through the
WITH GRANT OPTION on the GRANT statement. However, the WITH
GRANT OPTION does not allow the person granting the privilege to revoke
the privilege once granted. You must have SECADM authority,
ACCESSCTRL authority, or the CONTROL privilege to revoke an explicitly
granted privilege.
򐂰 Implicit
Implicit privileges are granted implicitly through the execution of a command.
For example, if a user executes a CREATE statement, they inherently gain
full access to the object they created, otherwise known as CONTROL
privileges.
db2 CREATE TABLE t2
In the example above, the user who issued the CREATE TABLE statement
have full CONTROL privileges over table t2.
Possessing the CONTROL privilege on an object allows a user to access that
database object, and to grant and revoke privileges to or from other users on
that object. The creator of a base table, materialized query table, staging
table, or nickname automatically receives the CONTROL privilege. The
creator of a view automatically receives the CONTROL privilege if they hold

186 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

the CONTROL privilege on all tables, views, and nicknames identified in the
fullselect.
򐂰 Indirect
When a user has the privilege to execute a package or routine, they do not
necessarily require specific access privileges on the objects handled in the
package or routine. If the package or routine contains static SQL or XQuery
statements, the privileges of the owner of the package are used for those
statements. If the package or routine contains dynamic SQL or XQuery
statements, the authorization ID used for privilege checking depends on the
setting of the DYNAMICRULES BIND option of the package issuing the
dynamic query statements, and whether those statements are issued when
the package is being used in the context of a routine.
A user or group can be authorized for any combination of individual privileges
or authorities. When a privilege is associated with a resource, that resource
must exist. For example, a user cannot be given the SELECT privilege on a
table unless that table has previously been created.

Note: Care must be given to granting authorities and privileges to user names
that do not exist in the system yet. At some later time, this user name can be
created and automatically receive all of the authorities and privileges
previously granted.

Privileges and authorities in DB2 can be granted on the following levels:

򐂰 Database level
– CONNECT privilege
– CREATETAB privilege
– LOAD privilege
– IMPLICIT_SCHEMA privilege
– BINDADD privilege
– CREATE_EXTERNAL_ROUTINE privilege
– CREATE_NOT_FENCED_ROUTINE privilege
– IMPLICIT_SCHEMA privilege
– LOAD privilege
– QUIESCE_CONNECT privilege
– ACCESSCTRL authority
– DATAACCESS authority
– EXPLAIN authority
– DBADM authority
– SECADM authority
– SQLADM authority
– WLMADM authority
– Schema level

Chapter 7. Data conversion 187

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

– CREATEIN privilege
– ALTERIN privilege
– DROPIN privilege
򐂰 Table space level
– USE privilege
򐂰 Table and view level
– CONTROL privilege
– SELECT privilege
– INSERT privilege
– UPDATE privilege
– DELETE privilege
– INDEX privilege
– ALTER privilege
– REFERENCES privilege
– ALL PRIVILEGES privilege
򐂰 Row or column level
– SECURITY LABEL privilege
– LBAC Rule Exemption privileges
򐂰 Other privileges
– Package privileges
– Index level
– Procedure, function and method privileges
– Sequence privileges

Label-Based Access Control (LBAC)

Label-based access control (LBAC) was introduced in DB2 9. LBAC provides the
ability to have more control over who can access specific data in tables and
views. LBAC provides particular privileges for users, groups or roles on one or
more columns and/or row(s) within a table.

LBAC controls access to table objects by attaching security labels to rows and
columns. Users attempting to access an object must have a certain security label
granted to them. Only users who have the proper labels when accessing the row
are allowed to retrieve the data. Other users do not receive any indication that
they could not retrieve the row. This form of security is different from normal
SELECT privileges, as users who attempt to access a table that they are not
allowed to access receive an SQL error messages.

All privileges and labels can be granted to users or groups with the GRANT
command, they can be revoked using the REVOKE command.

188 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

Note: Catalogs contain, among other things, statistics about data distribution
in a table, which are needed by the query optimizer to determine the best way
to execute a query. Users may indirectly gain knowledge about some data
values by accessing the catalogs. When using Label-Based Access Control
(LBAC) the new RESTRICT_ACCESS option can be added on the CREATE
DATABASE command to create a database where access to the catalogs is
not granted to PUBLIC. Access to the catalogs can then be granted on a
“need-to-know” basis.

Mapping the user information from MySQL to DB2

When converting from MySQL to DB2, you must port user privileges as well.

Table 7-1 shows the mapping of MySQL privileges to DB2 privileges, assuming
that different MySQL databases are mapped to different DB2 schemas. For
example, an INSERT privilege granted in MySQL on the global level means that
you have to grant the INSERT privilege on all existing tables in the DB2 database
to the specified user. If you create a new table in DB2, you have to grant the
INSERT privilege on this table to the user.

Table 7-1 Mapping of MySQL to DB2 privileges

MySQL MySQL DB2 privilege or authority DB2 scope
Privilege Scope

ALL global SYSADMIN authority and DBADM Instance

[PRIVILEGES] authority with DATAACCESS and
ACCESSCTRL authorities

database DBADM authority with Database

DATAACCESS and ACCESSCTRL
authorities

table ALL WITH GRANT OPTION Table

ALTER global ALTERIN For each schema

in the database

database ALTERIN Schema

table ALTER TABLE Table

ALTER global CREATE_EXTERNAL_ROUTINE Database
ROUTINE

database CREATIN Schema

CREATE global CREATETAB Database

database CREATIN Schema

table CONTROL Table

Chapter 7. Data conversion 189

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

CREATE global CREATE_EXTERNAL_ROUTINE Database

ROUTINE

database CREATIN Schema

CREATE global CREATETAB Database

TEMPORARY
TABLES

database CREATIN Schema

CREATE USER global SECADMIN authority Database

CREATE VIEW global DATAACCESS authority Database

database CONTROL or SELECT For all tables in the

Schema
table CONTROL or SELECT Table

DELETE global DATAACCESS authority Database

database DELETE For each table in

the schema

table DELETE Table

DROP global DROPIN For each schema

in the database

database DROPIN Schema

table CONTROL Table

EVENT global DBADM authority Database

database DBADM authority Database

EXECUTE global DATAACCESS authority Database

database EXECUTE For each routine in

the schema

routine EXECUTE Routine

FILE global LOAD authority Database

GRANT global SECADMIN authority or DBADM Database

OPTION WITH ACCESSCTRL

database SECADMIN authority or DBADM Database

WITH ACCESSCTRL

table SECADMIN authority or DBADM Database/Table

WITH ACCESSCTRL or CONTROL

routine SECADMIN authority or DBADM Routine

WITH ACCESSCTRL or CONTROL

INDEX global CREATETAB Database

database CREATIN Schema

190 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

table INDEX Table

INSERT global DATAACCESS authority Database

database INSERT For all tables in the

Schema

table INSERT Table

column LBAC Column

PROCESS global SYSADM, SYSCTRL, or Instance

SYSMAINT authority
RELOAD global WLMADM Database

REPLICATION global DBADM authority Database

CLIENT

REPLICATION global DBADM authority Database

SLAVE

SELECT global DATAACCESS authority Database

database SELECT For all tables in the

Schema

table SELECT Table

column LBAC Column

SHOW global ALL users All users
DATABASES

SHOW VIEW global SELECT on

SYSCAT.VIEWS

database SELECT on
SYSCAT.VIEWS

table SELECT on
SYSCAT.VIEWS

SHUTDOWN global SYSADMN, SYSCTRL or Instance

SYSMAINT authority

SUPER global Not available

TRIGGER global DBADM authority Database

database ALTERIN Schema

table CONTROL Table

UPDATE global DATAACCESS authority Database

database UPDATE For all tables in the

Schema

table UPDATE Table

column LBAC Column

Chapter 7. Data conversion 191

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

USAGE global not applicable since users are

created through the operating
system

database not applicable since users are

created through the operating
system

table not applicable since users are

created through the operating
system

7.2 Sample project: data porting

This section describes the steps that necessary to port data for our sample
conversion project.

7.2.1 Export user data from MySQL

We use the mysqlaccess utility to get the user names that have access to the
database inventory out of the MySQL database. See Example 7-11.

Example 7-11 shows retrieving users with access to the sample project
database.

Example 7-11 Retrieve users with access to the sample project database
>mysqlaccesss % % inventory -b -U root -P
mysqlaccess Version 2.06, 20 Dec 2000
……

Sele Inse Upda Dele Crea Drop Relo Shut Proc File Gran Refe Inde Alte Show Supe Crea
Lock Exec Repl Repl Crea Show Crea Alte Crea Even Trig Ssl_ Ssl_ X509 X509 Max_ Max_
Max_ Max_ | Host,User,DB
---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ----
---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ----
---- ---- + --------------------
N N N N N N N N N N N N N N N N N N N N N N N N N N N N ? ? ? ? 0 0 0 0 |
%,root,inventory
Y Y Y Y Y Y N N N N N Y Y Y Y N Y Y Y Y Y Y Y Y Y Y Y Y ? ? ? ? 0 0 0 0 |
%,inventAppUser,inventory
N N N N N N N N N N N N N N N N N N N N N N N N N N N N ? ? ? ? 0 0 0 0 |
%,user01,inventory
N N N N N N N N N N N N N N N N N N N N N N N N N N N N ? ? ? ? 0 0 0 0 |
%,user02,inventory
N N N N N N N N N N N N N N N N N N N N N N N N N N N N ? ? ? ? 0 0 0 0 |
%,ANY_NEW_USER,inventory

192 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y ? ? ? ? 0 0 0 0 |
localhost,root,inventory
Y Y Y Y Y Y N N N N N Y Y Y Y N Y Y Y Y Y Y Y Y Y Y Y Y ? ? ? ? 0 0 0 0 |
localhost,inventAppUser,inventory
N N N N N N N N N N N N N N N N N N N N N N N N N N N N ? ? ? ? 0 0 0 0 |
localhost,user01,inventory
N N N N N N N N N N N N N N N N N N N N N N N N N N N N ? ? ? ? 0 0 0 0 |
localhost,user02,inventory
N N N N N N N N N N N N N N N N N N N N N N N N N N N N ? ? ? ? 0 0 0 0 |
localhost,ANY_NEW_USER,inventory

7.2.2 Map MySQL user data to db2 user data

The root user is the database administrator and is not used by our sample
application; we do not want to port this user. The MySQL root user is similar to
the instance owner in DB2.

The user accounts user1, users2 and ANY_NEW_USER (which is similar to the
logical DB2 group PUBLIC) do not have any privileges on our sample database,
so we do not need to upgrade these users either.

The only remaining user is the user inventAppUser which is the account our
application uses to connect to the MySQL database and manipulate data.

We want to add a new group for this user called db2app1.

The user’s privileges are set for SELECT, INSERT, UPDATE, DELETE,
CREATE, and DROP on database level for inventory, so we map these
privileges to SELECT, INSERT, UPDATE and DELETE for all tables in the DB2
schema inventory and CREATEIN and DROPIN for the schema.

7.2.3 Create DB2 user

For user and group creation we use the script in Example 7-12.

Example 7-12 Sample script to create DB2 users and groups

db2server:~ # cat db2addusr.sh
export DB2DIR='/home/db2inst1' export HOMEDIR='/home'

groupadd $2
useradd -g $2 -m -d $HOMEDIR/$1 $1
passwd $1
echo '. '${DB2DIR}'/sqllib/db2profile' >> $HOMEDIR/$1/.profile

Chapter 7. Data conversion 193

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

The creation of our user and group was done by the root user as shown in
Example 7-13.

Example 7-13 Creation of the user and group

db2server:~ # ./db2addusr.sh inventAppUser db2app1
Changing password for inventAppUser.
New password:
Re-enter new password:
Password changed

The granting of the privileges was done by the instance owner db2inst1 with the
DB2 command shown in Example 7-14.

Example 7-14 Granting of privileges

db2 => connect to invent
connect to invent

Database Connection Information

Database server = DB2/LINUX 9.7.0

SQL authorization ID = DB2INST1
Local database alias = INVENT

grant createin, dropin on schema db2inst1 to group db2app1

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.inventory to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.locations to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.owners to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.services to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.BOSSGROUP to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.EMPGROUP to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.GENERALGROUP to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.GROUPS to user inventAppUSer

DB20000I The SQL command completed successfully.

194 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

grant select, insert, update, delete on table admin.MANAGERGROUP to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.SEVERITY to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.STATUS to user inventAppUSer

DB20000I The SQL command completed successfully.

grant select, insert, update, delete on table admin.TECHGROUP to user inventAppUSer

DB20000I The SQL command completed successfully.

7.2.4 Export MySQL application data

The database DDL scripts were generated and ported to DB2 automatically with
the IBM Data Movement Tool as described in 6.5, “Sample database conversion”
on page 148.

The next step is to extract the MySQL application data using the Data Movement
Tool. In 7.1.1, “Data porting commands and tools” on page 168 of this chapter we
discuss how to create data extraction and transfer files using the IBM Data
Movement Tool. And in 7.1.2, “Differences in data formats” on page 175 we
discuss how to modify the extraction scripts to extract special data. Now we must
execute these scripts and extract the data from the MySQL database. To do this,
open the Extract/Deploy window, un-select the DDL checkbox, select the Data
Movement checkbox and click Extract DDL/data, as shown in Figure 7-3 on
page 196. If you have made changes to the extraction script be sure to select No
when requested to recreate output directory or <tableName>.tables file.

Chapter 7. Data conversion 195

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

Figure 7-3 Deploy to Target Window

Now you can go into your project directory and check the extracted data files.
The data output files extracted from the MySQL database are located under the
<migration output directory>/data. Example 7-15 shows the created scripts for
our inventory scenario.

Example 7-15 IBM Data Movement Tool export files

db2server:/opt/ibm/IBMDataMovementTool/migr/data # ls -l
-rw-r--r-- 1 db2inst1 db2iadm1 114 Aug 28 01:03 admin_status.txt
-rw-r--r-- 1 db2inst1 db2iadm1 82 Aug 28 01:03 admin_severity.txt
-rw-r--r-- 1 db2inst1 db2iadm1 43188 Aug 28 01:03 admin_services.txt
-rw-r--r-- 1 db2inst1 db2iadm1 60137 Aug 28 01:03 admin_owners.txt
-rw-r--r-- 1 db2inst1 db2iadm1 4173 Aug 28 01:03 admin_locations.txt
-rw-r--r-- 1 db2inst1 db2iadm1 45489 Aug 28 01:03 admin_inventory.txt
-rw-r--r-- 1 db2inst1 db2iadm1 143 Aug 28 01:03 admin_groups.txt

Each of the files is tab-delimited, containing the data from the corresponding
MySQL table. This format can be read by the DB2 LOAD command.

196 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

7.2.5 Convert MySQL application data to DB2 format

As no special data types are used in our sample database, no manual conversion
has to be done in our export files.

7.2.6 Import application data into DB2

Importing data into DB2 can be done automatically by using the IBM Data
Movement Tool. The IBM Data Movement Tool creates a script containing the
DB2 LOAD commands and then executes the script. In case of any errors, the
script can be edited and run again.

Example 7-16 shows the DB2 LOAD commands generated by the IBM Data
Movement Tool for our sample project.

Example 7-16 DB2 LOAD commands for loading the data into the DB2 database
CONNECT TO INVENT;
--#SET :LOAD:ADMIN:GROUPS
LOAD FROM
"/opt/ibm/IBMDataMovementTool/migr/data/admin_groups.txt" OF DEL
MODIFIED BY CODEPAGE=1208COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR
NOROWWARNINGS --DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_groups.txt"
METHOD P(1,2,3,4,5)MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_groups.txt"
REPLACE INTO "ADMIN"."GROUPS"("GROUPNAME", "EDITUSER", "EDITGRANTUSERPERM", "EDITINVT",
"EDITSERVICE") --STATISTICS YES WITH DISTRIBUTION AND DETAILED INDEXES ALL
NONRECOVERABLE INDEXING MODE AUTOSELECT;

--#SET :LOAD:ADMIN:LOCATIONS
LOAD FROM
"/opt/ibm/IBMDataMovementTool/migr/data/admin_locations.txt" OF DEL
MODIFIED BY CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR
NOROWWARNINGS --DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_locations.txt"
METHOD P(1,2,3,4)MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_locations.txt"
REPLACE INTO "ADMIN"."LOCATIONS" ("ID", "ROOMNAME", "FLOORNUM", "PASSCODE")
--STATISTICS YES WITH DISTRIBUTION AND DETAILED INDEXES ALL NONRECOVERABLE
INDEXING MODE AUTOSELECT ;

--#SET :LOAD:ADMIN:SEVERITY
LOAD FROM
"/opt/ibm/IBMDataMovementTool/migr/data/admin_severity.txt" OF DEL
MODIFIED BY CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR
NOROWWARNINGS --DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_severity.txt"
METHOD P (1,2,3,4,5) MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_severity.txt"
REPLACE INTO "ADMIN"."SEVERITY" ("ID", "TITLE", "NOTES", "ESTDAYS", "AVGDAYS"
)--STATISTICS YES WITH DISTRIBUTION AND DETAILED INDEXES ALL NONRECOVERABLE INDEXING
MODE AUTOSELECT;

Chapter 7. Data conversion 197

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

--#SET :LOAD:ADMIN:OWNERS
LOAD FROM
"/opt/ibm/IBMDataMovementTool/migr/data/admin_owners.txt" OF DEL
MODIFIED BY CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR
NOROWWARNINGS --DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_owners.txt"
METHOD P (1,2,3,4,5,6,7,8,9,10,11) MESSAGES
"/opt/ibm/IBMDataMovementTool/migr/msg/admin_owners.txt" REPLACE INTO "ADMIN"."OWNERS"
("ID", "FIRSTNAME", "LASTNAME", "EMAIL", "LOCID", "CUBENUM",
"PHONENUM", "LOGINNAME", "PASSWORD", "FAXNUM", "GROUPS" ) --STATISTICS YES WITH
DISTRIBUTION AND DETAILED INDEXES ALL NONRECOVERABLE INDEXING MODE AUTOSELECT;

--#SET :LOAD:ADMIN:INVENTORY
LOAD FROM
"/opt/ibm/IBMDataMovementTool/migr/data/admin_inventory.txt" OF DEL
MODIFIED BY CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR
NOROWWARNINGS --DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_inventory.txt"
METHOD P (1,2,3,4,5,6,7,8) MESSAGES
"/opt/ibm/IBMDataMovementTool/migr/msg/admin_inventory.txt"
REPLACE INTO "ADMIN"."INVENTORY"("ID", "ITEMNAME", "MANUFACTURER", "MODEL", "YEAR",
"SERIAL", "LOCID", "OWNERID")--STATISTICS YES WITH DISTRIBUTION AND DETAILED INDEXES ALL
NONRECOVERABLE INDEXING MODE AUTOSELECT ;

--#SET :LOAD:ADMIN:SERVICES
LOAD FROM
"/opt/ibm/IBMDataMovementTool/migr/data/admin_services.txt" OF DEL
MODIFIED BY CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR
NOROWWARNINGS --DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_services.txt"
METHOD P (1,2,3,4,5,6,7,8,9) MESSAGES
/opt/ibm/IBMDataMovementTool/migr/msg/admin_services.txt" REPLACE INTO
"ADMIN"."SERVICES" ("ID", "INVENTID", "DESCRIPTION", "SEVERITY", "SERVICEOWNER",
"OPENDATE", "CLOSEDATE", "TARGETCLOSEDATE", "STATUS" )--STATISTICS YES WITH DISTRIBUTION
AND DETAILED INDEXES ALL NONRECOVERABLE INDEXING MODE AUTOSELECT;

--#SET :LOAD:ADMIN:STATUS
LOAD FROM
"/opt/ibm/IBMDataMovementTool/migr/data/admin_status.txt" OF DEL
MODIFIED BY CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR
NOROWWARNINGS --DUMPFILE="/opt/ibm/IBMDataMovementTool/migr/dump/admin_status.txt"
METHOD P (1,2,3) MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_status.txt"
REPLACE INTO "ADMIN"."STATUS" ("ID", "TITLE", "NOTES" )--STATISTICS YES WITH
DISTRIBUTION AND DETAILED INDEXES ALL NONRECOVERABLE INDEXING MODE AUTOSELECT;

TERMINATE;

To execute the DB2 LOAD commands go to the Extract/Deploy window. Make

sure that the DDL checkbox is still unselected and the Date Movement is
selected. Click the Deploy DDL/Data button, as shown in Figure 7-3 on

198 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

page 196. You can also execute this load from the command line by running the
db2load.sh script. If you have made changes to the extraction script be sure to
select No when requested to recreate output directory or <tableName>.tables
file, as it overrides your changes.

After importing the data into the DB2 tables, the RUNSTATS command should be
executed in order to recreate the statistic information about indexes. The
statistics information is used by the query optimizer. The IBM Data Movement
Tool generates a custom RUNSTATS script for the new database called
db2runstats.sql. Example 7-17 shows the db2runstats.sql script for our sample
conversion. You can run this in the IBM Data Movement Tool GUI or command
line.

Example 7-17 DB2 RUBSTATS commands for recreating the statistics information
CONNECT TO INVENT;
RUNSTATS ON TABLE "ADMIN"."GROUPS" ON ALL COLUMNS WITH DISTRIBUTION
ON ALL COLUMNS AND DETAILED INDEXES ALL ALLOW WRITE ACCESS ;

RUNSTATS ON TABLE "ADMIN"."LOCATIONS" ON ALL COLUMNS WITH DISTRIBUTION

ON ALL COLUMNS AND DETAILED INDEXES ALL ALLOW WRITE ACCESS ;

RUNSTATS ON TABLE "ADMIN"."SEVERITY" ON ALL COLUMNS WITH DISTRIBUTION

ON ALL COLUMNS AND DETAILED INDEXES ALL ALLOW WRITE ACCESS ;

RUNSTATS ON TABLE "ADMIN"."OWNERS" ON ALL COLUMNS WITH DISTRIBUTION

ON ALL COLUMNS AND DETAILED INDEXES ALL ALLOW WRITE ACCESS ;

RUNSTATS ON TABLE "ADMIN"."INVENTORY" ON ALL COLUMNS WITH DISTRIBUTION

ON ALL COLUMNS AND DETAILED INDEXES ALL ALLOW WRITE ACCESS ;

RUNSTATS ON TABLE "ADMIN"."SERVICES" ON ALL COLUMNS WITH DISTRIBUTION

ON ALL COLUMNS AND DETAILED INDEXES ALL ALLOW WRITE ACCESS ;

RUNSTATS ON TABLE "ADMIN"."STATUS" ON ALL COLUMNS WITH DISTRIBUTION

ON ALL COLUMNS AND DETAILED INDEXES ALL ALLOW WRITE ACCESS ;

7.2.7 Basic data checking

Once the script is executed, a db2load.log file is created where all output
messages are logged. Check the log file for the success of the DB2 LOAD
commands. You can find this information in the log file as shown in
Example 7-18.

Example 7-18 Log file information about the DB2 LOAD command
Number of rows read = 140
Number of rows skipped = 0
Number of rows loaded = 140
Number of rows rejected = 0

Chapter 7. Data conversion 199

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

Number of rows deleted = 0

Number of rows committed = 140

LOAD FROM "/opt/ibm/IBMDataMovementTool/migr/data/admin_severity.txt" OF DEL MODIFIED BY

CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR NOROWWARNINGS METHOD P
(1,2,3,4,5) MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_severity.txt" REPLACE INTO
"ADMIN"."SEVERITY" ( "ID", "TITLE", "NOTES", "ESTDAYS", "AVGDAYS" ) NONRECOVERABLE INDEXING MODE
AUTOSELECT

Number of rows read = 5

Number of rows skipped = 0
Number of rows loaded = 5
Number of rows rejected = 0
Number of rows deleted = 0
Number of rows committed = 5

LOAD FROM "/opt/ibm/IBMDataMovementTool/migr/data/admin_owners.txt" OF DEL MODIFIED BY

CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR NOROWWARNINGS METHOD P
(1,2,3,4,5,6,7,8,9,10,11) MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_owners.txt" REPLACE
INTO "ADMIN"."OWNERS" ( "ID", "FIRSTNAME", "LASTNAME", "EMAIL", "LOCID", "CUBENUM", "PHONENUM",
"LOGINNAME", "PASSWORD", "FAXNUM", "GROUPS" ) NONRECOVERABLE INDEXING MODE AUTOSELECT

Number of rows read = 504

Number of rows skipped = 0
Number of rows loaded = 502
Number of rows rejected = 2
Number of rows deleted = 0
Number of rows committed = 504

SQL3107W There is at least one warning message in the message file.

LOAD FROM "/opt/ibm/IBMDataMovementTool/migr/data/admin_inventory.txt" OF DEL MODIFIED BY

CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR NOROWWARNINGS METHOD P
(1,2,3,4,5,6,7,8) MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_inventory.txt" REPLACE INTO
"ADMIN"."INVENTORY" ( "ID", "ITEMNAME", "MANUFACTURER", "MODEL", "YEAR", "SERIAL", "LOCID", "OWNERID"
) NONRECOVERABLE INDEXING MODE AUTOSELECT

Number of rows read = 703

Number of rows skipped = 0
Number of rows loaded = 703
Number of rows rejected = 0
Number of rows deleted = 0
Number of rows committed = 703

SQL3107W There is at least one warning message in the message file.

LOAD FROM "/opt/ibm/IBMDataMovementTool/migr/data/admin_services.txt" OF DEL MODIFIED BY

CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR NOROWWARNINGS METHOD P
(1,2,3,4,5,6,7,8,9) MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_services.txt" REPLACE INTO
"ADMIN"."SERVICES" ( "ID", "INVENTID", "DESCRIPTION", "SEVERITY", "SERVICEOWNER", "OPENDATE",
"CLOSEDATE", "TARGETCLOSEDATE", "STATUS" ) NONRECOVERABLE INDEXING MODE AUTOSELECT

Number of rows read = 808

Number of rows skipped = 0
Number of rows loaded = 808
Number of rows rejected = 0
Number of rows deleted = 0
Number of rows committed = 808

LOAD FROM "/opt/ibm/IBMDataMovementTool/migr/data/admin_status.txt" OF DEL MODIFIED BY

CODEPAGE=1208 COLDEL~ ANYORDER USEDEFAULTS CHARDEL"" DELPRIORITYCHAR NOROWWARNINGS METHOD P
(1,2,3) MESSAGES "/opt/ibm/IBMDataMovementTool/migr/msg/admin_status.txt" REPLACE INTO
"ADMIN"."STATUS" ( "ID", "TITLE", "NOTES" ) NONRECOVERABLE INDEXING MODE AUTOSELECT

200 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

Number of rows read = 7

Number of rows skipped = 0
Number of rows loaded = 7
Number of rows rejected = 0
Number of rows deleted = 0
Number of rows committed = 7

TERMINATE
DB20000I The TERMINATE command completed successfully.

Database Connection Information

Database server = DB2/LINUX 9.7.0

SQL authorization ID = DB2INST1
Local database alias = INVENT

CONNECT TO INVENT

Database Connection Information

Database server = DB2/LINUX 9.7.0

SQL authorization ID = DB2INST1
Local database alias = INVENT

select count_big(*) "ADMIN.GROUPS" FROM "ADMIN"."GROUPS"

ADMIN.GROUPS
---------------------------------
6.

1 record(s) selected.

select count_big(*) "ADMIN.LOCATIONS" FROM "ADMIN"."LOCATIONS"

ADMIN.LOCATIONS
---------------------------------
140.

1 record(s) selected.

select count_big(*) "ADMIN.SEVERITY" FROM "ADMIN"."SEVERITY"

ADMIN.SEVERITY
---------------------------------
5.

1 record(s) selected.

select count_big(*) "ADMIN.OWNERS" FROM "ADMIN"."OWNERS"

ADMIN.OWNERS
---------------------------------
502.

1 record(s) selected.

select count_big(*) "ADMIN.INVENTORY" FROM "ADMIN"."INVENTORY"

ADMIN.INVENTORY

Chapter 7. Data conversion 201

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

---------------------------------
703.

1 record(s) selected.

select count_big(*) "ADMIN.SERVICES" FROM "ADMIN"."SERVICES"

ADMIN.SERVICES
---------------------------------
808.

1 record(s) selected.

select count_big(*) "ADMIN.STATUS" FROM "ADMIN"."STATUS"

ADMIN.STATUS
---------------------------------
7.

1 record(s) selected.

TERMINATE
DB20000I The TERMINATE command completed successfully.

Database Connection Information

Database server = DB2/LINUX 9.7.0

SQL authorization ID = DB2INST1
Local database alias = INVENT

TABLE_NAME FK_CHECKED CC_CHECKED STATUS

-------------------------------------------------- ---------- ---------- ------
ADMIN.GROUPS Y Y N

1 record(s) selected.

TABLE_NAME FK_CHECKED CC_CHECKED STATUS

-------------------------------------------------- ---------- ---------- ------
ADMIN.LOCATIONS Y Y N

1 record(s) selected.

TABLE_NAME FK_CHECKED CC_CHECKED STATUS

-------------------------------------------------- ---------- ---------- ------
ADMIN.SEVERITY Y Y N

1 record(s) selected.

TABLE_NAME FK_CHECKED CC_CHECKED STATUS

-------------------------------------------------- ---------- ---------- ------
ADMIN.OWNERS Y Y N

1 record(s) selected.

202 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

TABLE_NAME FK_CHECKED CC_CHECKED STATUS

-------------------------------------------------- ---------- ---------- ------
ADMIN.INVENTORY Y Y N

1 record(s) selected.

TABLE_NAME FK_CHECKED CC_CHECKED STATUS

-------------------------------------------------- ---------- ---------- ------
ADMIN.SERVICES Y Y N

1 record(s) selected.

TABLE_NAME FK_CHECKED CC_CHECKED STATUS

-------------------------------------------------- ---------- ---------- ------
ADMIN.STATUS Y Y N

1 record(s) selected.

DB20000I The TERMINATE command completed successfully.

Make sure that the number of rows read equals the number of rows committed.
This should also equal to the number of records in the MySQL source table.
Example 7-19 shows the MySQL command for the record count.

Example 7-19 Retrieving the number of records from MySQL

mysql> select count(*) from groups;
+----------+
| count(*) |
+----------+
| 6 |
+----------+
1 row in set (0.04 sec)

mysql> select count(*) from inventory;

+----------+
| count(*) |
+----------+
| 703 |
+----------+
1 row in set (0.00 sec)

mysql> select count(*) from locations;

+----------+
| count(*) |
+----------+
| 140 |
+----------+
1 row in set (0.00 sec)

mysql> select count(*) from owners;

+----------+
| count(*) |
+----------+
| 504 |
+----------+
1 row in set (0.01 sec)

Chapter 7. Data conversion 203

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

mysql> select count(*) from services;

+----------+
| count(*) |
+----------+
| 808 |
+----------+
1 row in set (0.00 sec)

mysql> select count(*) from status;

+----------+
| count(*) |
+----------+
| 7 |
+----------+
1 row in set (0.00 sec)

mysql> select count(*) from severity;

+----------+
| count(*) |
+----------+
| 5 |
+----------+
1 row in set (0.00 sec)

After you have checked that all the records were loaded into DB2, you should
check the sample data in each ported table. Pay attention to ensure that the
values are correct, especially if you have any-time values or decimal values.
Example 7-20 shows the table content checking.

Example 7-20 Sample data of MySQL data

mysql> select * from services where id = 20;
+----+----------+-------------+----------+--------------+------------+------------+-----
------------+--------+
| id | inventID | description | severity | serviceOwner | openDate | closeDate |
targetCloseDate | status |
+----+----------+-------------+----------+--------------+------------+------------+-----
------------+--------+
| 20 | 36 | test | 1 | 108 | 2009-01-18 | 2009-01-20 |
2009-01-19 | 7 |
+----+----------+-------------+----------+--------------+------------+------------+-----
------------+--------+
1 row in set (0.00 sec)

db2inst1@db2server:~> db2 "select * from admin.services where id = 20"

ID INVENTID DESCRIPTION SEVERITY SERVICEOWNER OPENDATE CLOSEDATE TARGETCLOSEDATE

STATUS
----------- ----------- -------------------- ----------- ------------ ----------
---------- --------------- -----------
20 36 test 1 108 2009-01-18 2009-01-20 2009-01-19
7

1 record(s) selected.

204 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch07.fm

Chapter 7. Data conversion 205

7093ch07.fm Draft Document for Review October 16, 2009 10:21 am

206 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Chapter 8. Application conversion

The task of converting an application, its databases, and the associated data
most often requires significant resources and commitment. Simultaneously to the
meticulous planning of the porting project as a whole, it is paramount to assess
the issues that may be high up to the highest level of resources. In many porting
projects this is the part of application porting. This chapter attempts to give the
reader “food for thought” in the following areas:
򐂰 Differences in SQL Data Manipulation Language (DML), built-in functions,
and SQL semantics
򐂰 Converting application source, application programming interfaces (APIs),
and condition handling
򐂰 Internals of the Database Management System that may impact the
conversion, such as locking, isolation levels, transaction logging, and national
language support

© Copyright IBM Corp. 2009. All rights reserved. 207

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

8.1 Data Manipulation Language differences and

similarities
Modifying an application to work with DB2 is an important task during the
conversion process. While a large portion of this step may encompass changing
code to work with a different development environment, it is likely that additional
time will be spent on testing the resulting code.

8.1.1 SELECT syntax

This section focuses on the SELECT statement syntax as it is supported in
MySQL, and attempts to show how MySQL extensions to the SQL standard
might be implemented in DB2. Example 8-1 shows the MySQL syntax for the
SELECT statement. A discussion of individual keywords follows.

Example 8-1 MySQL SELECT syntax

SELECT [STRAIGHT_JOIN]
[SQL_SMALL_RESULT] [SQL_BIG_RESULT] [SQL_BUFFER_RESULT]
[SQL_CACHE | SQL_NO_CACHE] [SQL_CALC_FOUND_ROWS] [HIGH_PRIORITY]
[DISTINCT | DISTINCTROW | ALL]
select_expression,...
[INTO {OUTFILE | DUMPFILE} 'file_name' export_options]
[FROM table_references
[WHERE where_definition]
[GROUP BY {unsigned_integer | col_name | formula} [ASC | DESC], ... [WITH
ROLLUP]]
[HAVING where_definition]
[ORDER BY {unsigned_integer | col_name | formula} [ASC | DESC], ...] [LIMIT
[offset,] row_count | row_count OFFSET offset]
[PROCEDURE procedure_name(argument_list)]
[FOR UPDATE | LOCK IN SHARE MODE]]

The STRAIGHT_JOIN keyword forces the MySQL optimizer to join tables in the
order specified. In DB2 the join order is always determined by the optimizer. The
optimizer choices can be limited by changing the default query optimization class
to a lower level using SET CURRENT QUERY OPTIMIZATION. This however,
does not guarantee that the optimizer will evaluate the join in the order stated
within the SQL statement since the DB2 cost-based optimizer usually chooses
the best access path for a given query. For additional information see 10.5.5,
“SQL execution plan” on page 365.

Options prefixed with SQL are MySQL specific and do not require a DB2
equivalent.

208 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

The DISTINCTROW keyword is a synonym for the DISTINCT clause supported

by DB2.

The INTO {OUTFILE | DUMPFILE} 'file_name' export_options selection allows

you to write output data to a file without invoking the mysqldump utility. The DB2
command line processor allows you to direct the output of any SELECT
statement to an operating system file.

The LIMIT [offset,] row_count | row_count OFFSET offset keyword translates to

FETCH FIRST n ROWS ONLY in DB2. An offset to retrieve rows needs to be
implemented through the WHERE clause if possible.

With the SQL_SMALL_RESULT or SQL_BIG_RESULT commands, the query

developer can hint to the SQL optimizer the size of the expected result set,
influencing the optimizer access strategy. Example 8-2 shows how the optimizer
hint works.

Example 8-2 MySQL SELECT with optimizer hint

mysql> SELECT sql_small_result * FROM t1;

DB2 has a similar operator to guide SQL optimizer decisions with a different
syntax as shown in Example 8-3.

Example 8-3 DB2 SELECT with optimizer hint

db2 => SELECT * FROM t1 OPTIMIZE FOR 2 ROWS
DB20000I The SQL command completed successfully.

8.1.2 JOIN syntax

Join capabilities for commercial and industrial strength database management
systems are very important. MySQL supports the linguistic elements for JOIN as
shown in Example 8-4. Various aspects of JOINs are discussed below.

Example 8-4 MySQL JOIN Syntax

table_reference, table_reference
table_reference [INNER | CROSS] JOIN table_reference [join_condition]
table_reference STRAIGHT_JOIN table_reference
table_reference LEFT [OUTER] JOIN table_reference [join_condition]
table_reference NATURAL [LEFT [OUTER]] JOIN table_reference
{OJ table_reference LEFT OUTER JOIN table_reference ON conditional_expr }

table_reference RIGHT [OUTER] JOIN table_reference [join_condition]

table_reference NATURAL [RIGHT [OUTER]] JOIN table_reference

Chapter 8. Application conversion 209

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

Where table_reference is defined as:

table_name [[AS] alias] [[USE INDEX (key_list)] | [IGNORE INDEX (key_list)] |
[FORCE IN and join_condition is defined as:
ON conditional_expr |
USING (column_list)

The STRAIGHT_JOIN keywords force the MySQL optimizer to join tables in the
order they are specified. In DB2 the join order is always determined by the
optimizer. The optimizer choices can be limited by changing the default query
optimization class using SET CURRENT QUERY OPTIMIZATION command.

A NATURAL join, as its name implies, can be invoked when two or more tables
share exactly the same columns needed for a successful equijoin. It is
semantically equivalent to DB2 INNER JOIN or LEFT OUTER JOIN with the
respective join criteria specified in the ON clause.

According to the SQL ANSI standard when you must join tables that share more
than one column naturally, the JOIN ... USING syntax must be used. An
equivalent join can be composed using the DB2 supported join syntax in the ON
clause.

Cartesian products do happen from time to time, usually as the result of an

equijoin condition that has been missed in a query using DB2 syntax, even
though individuals dealing with development of applications and database
queries spend a lot of time trying to avoid them. However, one of the advantages
of the CROSS JOIN syntax is that a specific keyword is required to create a
Cartesian product. Therefore, when the CROSS JOIN syntax is used in your
conversion project, you can code a regular join in DB2 with a no join condition in
the WHERE clause.

8.1.3 UNION syntax

MySQL support for the UNION feature (shown in Example 8-5) is very similar to
the DB2 syntax, and therefore does not require additional discussion.

Example 8-5 UNION syntax in MySQL and DB2

SELECT ...
UNION [ALL | DISTINCT] SELECT ...
[UNION [ALL | DISTINCT] SELECT ...]

210 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

8.1.4 Subquery syntax

A subquery is a SELECT statement within another query or statement.
Therefore, subqueries can be found in other SELECT statements - either in the
column list, the WHERE clause, or the HAVING clause, and in INSERT,
UPDATE, and DELETE statements. Example 8-6 shows a sequence of
subqueries in a DELETE statement.

Example 8-6 Example for subqueries an a DELETE statement

DELETE FROM t1
WHERE col1 > ANY
(SELECT COUNT(*) FROM t2
WHERE NOT EXISTS
(SELECT col3 FROM t3
WHERE col3 =
(SELECT col4 FROM t4 UNION SELECT 1 FROM
(SELECT col5 ROM t5) AS t5 )));

MySQL provides similar subquery implementation to that of DB2. Nothing has

been found causing serious issues when moving from MySQL to DB2.

8.1.5 Grouping, having, and ordering

All ANSI SQL 92 standard grouping functions available in MySQL Version 5.1
are also available in DB2. In general, no significant differences are found in
areas of grouping, having, and ordering. However, beyond the SQL 92 standard
DB2 provides some interesting functionality that may enhance the ported
application significantly. Refer to DB2 manual SQL Reference, Volume 1,
SC10-4249. Table 8-1 contains a partial list the differences in two databases and
provides conversion examples. For a full list of MySQL grouping functions and
the DB2 equivalent, refer to A.1, “Grouping related functions” on page 398.

Table 8-1 MySQL and DB2 grouping related function

MySQL function MYSQL example DB2 function DB2 example Notes

AVG([DISTINCT] mysql> SELECT a, AVG ([DISTINCT | db2 " SELECT a, Returns the
expression) AVG(b) ALL] expression) AVG(b) average set of
FROM t1 FROM t1 numbers
GROUP BY a GROUP BY a"

Chapter 8. Application conversion 211

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

COUNT([DISTINCT] mysql> SELECT a, COUNT([DISTINCT| db2 " SELECT a, Returns the

expression, COUNT(b) ALL] expression). COUNT(b) number of rows
expression,...) FROM t1 FROM t1 or values in a set
GROUP BY a DB2 allows only one GROUP BY a" of rows or values.
expression:
Use CONCAT for
character data type
or
CHAR and CONCAT
on
numeric data types
MAX ([DISTINCT] mysql> SELECT a, MAX ([DISTINCT | db2 "SELECT a, Returns the
expression) MAX(b) ALL] expression) MAX(b) maximum value
FROM t1 FROM t1 in a set of values.
GROUP BY a GROUP BY a"

SUM([DISTINCT] mysql> SELECT a, SUM([DISTINCT | db2 " SELECT a, Returns the sum
expression) SUM(b) ALL] expression) sum(b) of a set of
FROM t1 FROM t1 numbers.
GROUP BY a GROUP BY a"

GROUP BY on alias mysql> SELECT a as x use column name for db2 " SELECT a Groups data by
FROM a grouping FROM t1 the alias name
GROUP BY x; GROUP BY a" provided.

GROUP BY on mysql> SELECT a use column name for db2 " SELECT a Groups data by
position FROM t1 grouping FROM t1 the position
GROUP BY 1 GROUP BY a" provided.

HAVING on alias mysql> SELECT a as x use column name in db2 " SELECT a Groups data
FROM t1 having clause FROM t1 meeting the
GROUP BY a GROUP BY a HAVING
HAVING x > 0 HAVING a > 0” expression.

8.1.6 Strings
Unless starting MySQL in ANSI mode (using mysqld --ansi), it behaves
differently than DB2. As Example 8-7 illustrates, MySQL accepts single as well
as double quotes as a string delimiter when started in default mode.

Example 8-7 MySQL string handling

mysql> select 'redbook', '"redbook"', '""redbook""', 'red''book';
+---------+-----------+-------------+----------+
| redbook| "redbook"| ""redbook""| red'book|
+---------+-----------+-------------+----------+
| redbook| "redbook"| ""redbook""| red'book|
+---------+-----------+-------------+----------+
1 row in set (0.00 sec)
mysql> select "redbook", "'redbook'", "''redbook''", "red""book";
+---------+-----------+-------------+----------+
| redbook | 'redbook'| ''redbook''| red"book|

212 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

+---------+-----------+-------------+----------+
| redbook | 'redbook' | ''redbook'' | red"book |
+----------+-------------+---------------+----------+
1 row in set (0.00 sec)

DB2 is designed and implemented according to the ANSI standard and therefore
accepts single quotes as a string delimiter. Double quotes are used in DB2 for
delimiting SQL identifiers. Example 8-8 shows how DB2 handles strings. Similar
results will be achieved when MySQL runs in ANSI mode.

Example 8-8 DB2 string handling

db2 => select "redbook", "'redbook'", "''redbook''", "red""book" from t1
SQL0206N "redbook" is not valid in the context where it is used.
SQLSTATE=42703

db2 => select 'redbook', '"redbook"', '""redbook""', 'red''book' from t1

1 2 3 4
---------------------------------
redbook "redbook" ""redbook"" red'book

Table 8-2 provides an overview of a few of the MySQL string related functions,
and how these can be converted to DB2. For a full list of MySQL string functions
and the DB2 equivalent, refer to A.2, “String functions” on page 400.

Table 8-2 MySQL and DB2 string related function

MySQL MYSQL example DB2 function DB2 example Notes
function

ASCII(string) mysql> SELECT ascii('a'); ASCII(string) db2 "VALUES ascii('a') " Returns ASCII
+------------+ code value
| ascii('a') | 1
+------------+ -----------
| 97 | 97
+------------+
1 row in set (0.00 sec) 1 record(s) selected

Chapter 8. Application conversion 213

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

CHAR_LEN mysql> SELECT CHARACTER db2 " VALUES Returns the

GTH(string) / CHAR_LENGTH('Orange'); _LENGTH(stri CHAR_LENGTH('Orang number of
CHARACTE +-----------------------+ ng, e', CODEUNITS32)" bytes for
R_LENGTH( | CHAR_LENGTH('Orange') | CODEUNITS expression. For
string) +-----------------------+ 16| 1 double byte
| 6| CODEUNITS ----------- character set
+-----------------------+ 32| 6 (DBCS) the
1 row in set (0.00 sec) OCTETS),/ number of
CHAR_LENG 1 record(s) selected. DBCS characters
TH(string, is
CODEUNITS returned
16|
CODEUNITS
32| OCTETS)
CONCAT_W mysql> SELECT Use || to db2 "SELECT (firstName Returns the
S(separator, CONCAT_WS('-', firstname, implement || '-' || lastName || '-' || concatenation of
string, lastname, loginname ) as CONCAT(list) loginName) as fullName string arguments
string,…) FULLNAME from owners from admin.owners with separator
where id = 501; where id = 501"
+-------------------------+
| FULLNAME | FULLNAME
+-------------------------+ ---------------------------------
| Angela-Carlson-acarlson | -------------------------
+-------------------------+ Angela-Carlson-acarlson
1 row in set (0.01 sec)
1 record(s) selected.

FORMAT(do FORMAT: No equivalent Refer to UDF B.2, Returns the

uble, integer) select function. “Sample code for rounded string
format(1234.5555, 2) Implement FORMAT function” on
returns 1,234.56 using UDF page 413
INSERT(strin mysql> SELECT INSERT(strin db2 "VALUES Inserts a string
g, position, INSERT('original', 2, 2, 'NEW'); g, position, INSERT('original', 2, 2, into an existing
length, +---------------------------------+ length, 'NEW')" string
substring) | INSERT('original', 2, 2, substring) 1
'NEW') | -----------
+---------------------------------+ oNEWginal
| oNEWginal | 1 record(s) selected.
+---------------------------------+
1 row in set (0.00 sec)
LPAD(string, mysql> SELECT No equivalent Refer to UDF B.3, Returns the a
length, LPAD('TEST',6,'!!!'); function. “Sample code for RPAD string of the given
substring) / +----------------------+ Implement and LPAD functions” on length, if the
RDAP( string, | LPAD('TEST',6,'!!!') | using UDF page 414 length is longer
length, +----------------------+ then the string the
substring) | !!TEST | substring
+----------------------+ characters will be
1 row in set (0.00 sec) added to the left or
right end.

214 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

REPLACE(st mysql> SELECT REPLACE REPLACE(stri db2 "VALUES REPLACE Returns as sting
ring1, string2, ('DINING', 'N', 'VID'); ng1, string2, ('DINING', 'N', 'VID') " with all
string3) +--------------------------------+ string3) 1 occurrences of
| REPLACE ('DINING', 'N', ------------------ string2 in string1
'VID') | DIVIDIVIDG with string3
+--------------------------------+ 1 record(s) selected.
| DIVIDIVIDG |
+--------------------------------+
1 row in set (0.00 sec)

SUBSTRING mysql> select SUBSTR(strin db2 "VALUES('abcdef', Returns a

(string, substring('abcdef', 2, 3); g, position, 2, 3)” substring of a
position, +---------------------------+ length) 1 string.
length) / MID | substring('abcdef', 2, 3) | ---
(string, +---------------------------+ bcd
position, | bcd |
length) / +---------------------------+ 1 record(s) selected.
SUBSTR(stri 1 row in set (0.00 sec)
ng, position,
[length])

TRIM([Both | mysql> select trim(trailing from TRIM([Both | db2 "VALUES Removes blanks
Leading | trim(LEADING FROM ' abc Leading | trim(trailing from or occurrences of
trailing ')) as OUTPUT; trailing trim(LEADING FROM ' another specified
[substring] +--------+ [substring] abc '))" character from the
FROM] | OUTPUT | FROM] string) end or the
string) +--------+ OUTPUT beginning of a
| abc | --------- string expression
+--------+ abc
1 row in set (0.00 sec)
1 record(s) selected.

8.1.7 Implicit casting of data types

DB2 9.7 introduces support for implicit casting. Implicit casting is an automatic
conversion of data from one data type to another based on an implied set of
conversion rules which occurs in support of weak typing.

Prior to Version 9.7, strong typing was used during comparisons and
assignments. Strong typing requires matching data types, which means that you
must explicitly convert one or both data types to a common data type prior too
performing comparisons or assignments.

In Version 9.7, the rules used during comparisons and assignments have been
relaxed. If reasonable interpretation can be made between two mismatched data
types, implicit casting is used to perform comparisons or assignments. Implicit
casting is also supported during function resolution. When the data types of the
arguments of a function being invoked cannot be promoted to the data types of
the parameters of the selected function, the data types of the arguments are
implicitly casted to those of the parameters.

Chapter 8. Application conversion 215

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

Implicit casting reduces the amount of SQL statements that you must modify
when enabling applications that run on data servers other than DB2 data servers
to run on DB2 9.7. In many cases, you no longer have to explicitly cast data
types when comparing or assigning values with mismatched data types.

Example 8-9 shows how MySQL implicitly casts the character value 5 to an
integer value to resolve the query.

Example 8-9 MySQL performs implicit data type casting

mysql> create table t1 (c1 int);
Query OK, 0 rows affected (0.08 sec)

mysql> insert into t1 values(5);

Query OK, 1 row affected (0.02 sec)

mysql> select * from t1 where c1='5';

+------+
| c1 |
+------+
| 5 |
+------+
1 row in set (0.00 sec)

An implicit casting of incompatible data types is now supported by DB2 as of DB2

9.7. Example 8-10 shows implicit casting of a character value in DB2 9.7.

Example 8-10 DB2 9.7 implicit data type casting

db2 => create table t1 (c1 int)
DB20000I The SQL command completed successfully.
db2 => insert into t1 values(5)
DB20000I The SQL command completed successfully.
db2 => commit
DB20000I The SQL command completed successfully.
db2 => select * from t1 where c1='5'

C1
-----------
5

1 record(s) selected.

For DB2 versions prior to 9.7 explicit casting of the character value to an integer
value is required as illustrated in Example 8-11.

Example 8-11 DB2 9.5 and prior requires explicit casting

db2 => select * from t1 where c1 = '5'

216 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

SQL0401N The data types of the operands for the operation "=" are not
compatible. SQLSTATE=42818
db2 => select * from t1 where c1 = cast ('5' as int)

C1
--------
5
1 record(s) selected.

Example 8-12 illustrates how MySQL implicitly casts numeric values and DATA,
TIME, or TIMESTAMP values to strings when concatenated.

Example 8-12 MySQL implicit casting using concatenation for strings and DATE
mysql> select concat('ITSOSJ',1234) from t1;
+-----------+
| stringcol |
+-----------+
| ITSOSJ1234|
+-----------+
1 row in set (0.02 sec)

mysql> select concat('ITSOSJ',current_date) as stringdate from t1;

+------------------+
| stringdate |
+------------------+
| ITSOSJ2009-08-31 |
+------------------+
1 row in set (0.01 sec)

Example 8-13 illustrates how DB2 9.7 implicitly casts numeric values, as well as
DATA, TIME, or TIMESTAMP values to strings when concatenated.

Example 8-13 DB2 9.7 casting character strings and DATE explicitly
db2 => select concat('ITSOSJ',1234) from t1

1
----------
ITSOSJ1234

1 record(s) selected.

db2 => select concat('ITSOSJ',current_date) as stringdate from t1

STRINGDATE
----------------
ITSOSJ08/31/2009

Chapter 8. Application conversion 217

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

1 record(s) selected.

DB2 9.5 and prior, require compatible arguments for the concatenation built-in
functions as shown in Example 8-14. If the arguments are not compatible, for
example calling a function with a character data type argument using a numeric
data type, the concatenation will fail with the error message SQL0440.

Example 8-14 DB2 9.5 and prior - casting character strings and DATE explicitly
db2 => select concat('ITSOSJ',1234) from t1
SQL0440N No authorized routine named "CONCAT" of type "FUNCTION" having
compatible arguments was found. SQLSTATE=42884
db2 => select concat('ITSOSJ','1234') as stringcol from t1

STRINGCOL
-----------
ITSOSJ1234

1 record(s) selected.

db2 => select concat('ITSOSJ', current date) as stringdate from t1

SQL0440N No authorized routine named "CONCAT" of type "FUNCTION" having
compatible arguments was found. SQLSTATE=42884

db2 => select concat('ITSOSJ',CAST(current date as char(20))) as stringdate

from t1

STRINGDATE
------------------------
ITSOSJ01/23/2004
1 record(s) selected.

8.1.8 String concatenation and NULL values

The ANSI92 standard for concatenation of multiple strings is stated to be
executed by the SQL || operator. MySQL does not support the || operator as
concatenation as it is used to represent an OR operator. MySQL uses the
CONCAT(string1, string2, string3, string4,….) function to concatenate strings, as
shown in Example 8-15.

Example 8-15 MySQL concatenation of strings

mysql> SELECT CONCAT('This ', 'is ', 'an ', 'example.') ;
+-------------------------------------------+
| CONCAT('This ', 'is ', 'an ', 'example.') |
+-------------------------------------------+

218 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

| This is an example. |
+-------------------------------------------+
1 row in set (0.00 sec)

DB2 follows the ANSI92 standard for concatenation of multiple strings. DB2 also
has a CONCAT(string1, string2) which can be used for concatenation of two
strings. Example 8-16 shows how DB2 handles concatenating strings.

Example 8-16 DB2 concatenation of strings

db2 => VALUES CONCAT('This ', 'is ', 'an ', 'example.')
SQL0440N No authorized routine named "CONCAT" of type "FUNCTION" having
compatible arguments was found. SQLSTATE=42884
db2 => VALUES CONCAT('This is ', 'an example.')

1
-------------------
This is an example.

1 record(s) selected.

db2 => VALUES ('This ' || 'is ' || 'an ' || 'example.')

1
-------------------
This is an example.

1 record(s) selected.

The ANSI92 standard states that if you concatenate a NULL value onto an
existing string, the result set is NULL. Example 8-17 shows you the behavior of
MySQL.

Example 8-17 MySQL concatenation of strings and NULL values

mysql> create table t2 (col1 char(2));
Query OK, 0 rows affected (0.03 sec)

mysql> insert into t2 values(NULL);

Query OK, 1 row affected (0.07 sec)
mysql> select concat('abc',col1) as nullstring from t2;
+-------------------+
| nullstring |
+-------------------+
| NULL |
+-------------------+
1 row in set (0.05 sec)
mysql> select concat('abc', coalesce(col1,'')) as nullstring from t2;

Chapter 8. Application conversion 219

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

+----------------------------------+
| nullstring |
+----------------------------------+
| abc |
+----------------------------------+ 1 row in set (0.00 sec)

As shown, MySQL behaves as ANSI-92 compliant, and therefore gives you the
same result sets as Example 8-18 for DB2.

Example 8-18 DB2 string and NuLL concatenation

db2 => create table t2 (col1 char(2))
DB20000I The SQL command completed successfully.
db2 => insert into t2 values(NULL)
DB20000I The SQL command completed successfully.
db2 => select concat('abc', col1) as nullstring from t2

NULLSTRING
-----
-
1 record(s) selected.

db2 => select concat('abc', coalesce(col1,'')) as nullstring from t2

NULLSTRING
-----
abc

1 record(s) selected.

8.1.9 Record deletion

Much like other competitive relational database management systems, MySQL
uses the TRUNCATE statement. TRUNCATE is used to delete all rows from a
table when there is no need to recover the deleted records. Therefore, there is no
ROLLBACK where a TRUNCATE is possible. Example 8-19 shows how
TRUNCATE is used to remove all rows, and does not allow a rollback of the
deleted rows.

Example 8-19 Using MySQL TRUNCATE to delete all records in a file

mysql> select * from t1;
+------+
| col1 |
+------+
| 5 |
| 10 |

220 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

+------+
2 rows in set (0.02 sec)

mysql> truncate table t1;

Query OK, 0 rows affected (0.00 sec)

mysql> select * from t1;

Empty set (0.00 sec)

mysql> rollback ;
ERROR 1196: Warning: Some non-transactional changed tables couldn't be rolled
back
mysql> select * from t1;
Empty set (0.00 sec)

The TRUNCATE option is primarily used to quickly delete all records from a table
when no recovery of the deleted rows is required. As of DB2 9.5 you can enable
the support of the TRUNCATE statement using the
DB2_COMPATIBILITY_VECTOR registry variable.

The DB2_COMPATIBILITY_VECTOR registry variable is used to enable one or

more DB2 compatibility features introduced since DB2 Version 9.5.

These features ease the task of converting applications written for other
relational database vendors to DB2 Version 9.5 or later. This DB2 registry
variable is represented as a hexadecimal value, and each bit in the variable
enables one of the DB2 compatibility features. To enable the TRUNCATE
statement set DB2_COMPATIBILITY_VECTOR registry variable to 8.
Example 8-20 shows the syntax to set the DB2_COMPATIBILITY_VECTOR
registry variable and execute the TRUNCATE command. More information on the
DB2_COMPATIBILITY_VECTOR registry variable can be found at the IBM
Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

Example 8-20 DB2 TRUNCATE to delete all records in a file

db2inst1@db2rules:~> db2stop force
08/31/2009 16:09:09 0 0 SQL1064N DB2STOP processing was successful.
SQL1064N DB2STOP processing was successful.

db2inst1@db2rules:~> db2set DB2_COMPATIBILITY_VECTOR=8

db2inst1@db2rules:~> db2set
DB2_COMPATIBILITY_VECTOR=8
DB2RSHCMD=/usr/bin/ssh
DB2COMM=tcpip

db2inst1@db2rules:~> db2start

Chapter 8. Application conversion 221

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

08/31/2009 16:13:19 0 0 SQL1063N DB2START processing was successful.

SQL1063N DB2START processing was successful.

db2inst1@db2rules:~> db2 create db itsodb

DB20000I The CREATE DATABASE command completed successfully.

db2inst1@db2rules:~> db2 connect to itsodb

Database Connection Information

Database server = DB2/LINUX 9.7.0

SQL authorization ID = DB2INST1
Local database alias = ISSODB

db2inst1@db2rules:~> db2 "create table t1 (c1 int) "

DB20000I The SQL command completed successfully.

db2inst1@db2rules:~> db2 "insert into t1 values(5)"

DB20000I The SQL command completed successfully.
db2inst1@db2rules:~> db2 "insert into t1 values(10)"
DB20000I The SQL command completed successfully.
db2inst1@db2rules:~> db2 "insert into t1 values(15)"
DB20000I The SQL command completed successfully.
db2inst1@db2rules:~> db2 "insert into t1 values(20)"
DB20000I The SQL command completed successfully.

db2inst1@db2rules:~> db2 "SELECT * from t1"

C1
-----------
5
10
15
20

4 record(s) selected.

db2inst1@db2rules:~> db2 TRUNCATE TABLE t1 IMMEDIATE

DB20000I The SQL command completed successfully.
db2inst1@db2rules:~> db2 "SELECT * from t1"

C1
-----------

0 record(s) selected.

You may also turn off logging with the following ALTER TABLE statement, to
achieve a similar behavior.
ALTER TABLE <tablename> ACTIVATE NOT LOGGED INITIALLY WITH EMPTY TABLE

222 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

8.1.10 Built-in functions and operators

We have already discussed string and aggregate functions. In this section we
complete the discussion of built-in functions, including date/time and numeric
functions. We also discuss the differences between various operators.

A function is an operation that is denoted by a function name followed by a pair of

parentheses enclosing the specification of arguments, if arguments are required.

The following group of tables is not an exhaustive list of differences in built-in

functions and operators between MySQL Version 5.1 and DB2 9.7. The listed
built-in functions attempt to highlight differences and mapping functionalities
between the two database management systems.

Numeric functions comparison

Many of the MySQL built-in numeric functions port easily to DB2 functions. A.3,
“Numeric functions” on page 406 lists some functions and examples specific to
manipulating numbers.

Functions converting date and time

Working with date and time functions and operators reveals many cases where
the implementation differences exist between the various DBMS. A.4, “Date and
time functions” on page 407 lists some date and time related functions and
examples.

Comparing operators and more functions

There is a plethora of functions as listed in A.5, “Comparing operators and other
functions” on page 408 where syntax or implementation can be quite different.
We summarize these differences and display code snippets for several selected
functions and ways to convert these from MySQL to DB2. In many cases you will
find the DB2 CASE expression is really helpful when function mapping is
required.

8.2 Application source conversion

Every programming language provides various commands and possibilities to
program a single functionality. It is almost impossible to give a conversion
example for every possible programming approach. Therefore, in this section we
discuss the most important concepts used to access a database, which are:
򐂰 connecting to a database
򐂰 query statements

Chapter 8. Application conversion 223

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 disconnecting from a database

8.2.1 Converting MySQL Perl applications to DB2

Perl uses the DBD::mysql interface driver as a plug-in for the Database
Independent (DBI) set of modules. DBI provides a common Perl API for all
database connections. For more information about the DBD::mysql driver visit:
http://search.cpan.org/~capttofu/DBD-mysql-4.012/lib/DBD/mysql.pm

The DBI interface can also be used to access DB2 using the DBD::DB2 driver.
Information about the DBI interface and the DBD::DB2 driver, as well as
installation instructions can be found at the following Web sites:
http://www.ibm.com/software/data/db2/perl/
http://search.cpan.org/~ibmtordb2/DBD-DB2-1.74/DB2.pod
http://www.perl.com/CPAN/modules/by-module/DBD/
http://aspn.activestate.com/ASPN/Modules/

Figure 8-1 on page 224 illustrates at how the Perl interfaces and pluggable
drivers connect to MySQL and DB2 databases.

Figure 8-1 Perl interface and pluggable drivers

Converting DBI interface supported code

Applications using the DBI interface to connect to MySQL generally can be
adapted to DB2 by simply changing the database driver from DBD::mysql to
DBD::DB2 within the code.

Connecting to a database
Use the following code within your Perl application to connect to a MySQL
database:
$dsn= "dbi:mysql:$database:$host:$port";
$dbh = DBI ->Connect($dsn, $user, $password);

224 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

The connection statement consists of the data source name, user ID, and
password. The data source name consists of the vendor specific database
driver, the database name, the host name and the port. Optionally, the host
name and port can be left out of the data source name. Example 8-21 on
page 225 shows the connect syntax to a MySQL database. For simplicity
reasons the error handling is not included in following examples.

Example 8-21 MySQL database connection with the Mysql.pm interface

use DBI;
use DBD::mysql;
my $host="localhost";
my $port="3306";
my $database="inventory";
my $user="mysql";
my $password="password";

my $dns="DBI:mysql:database=$database;host=$host;port=$port";

$dbh =
DBI->connect("$dns","$user","$password");

Connection to a DB2 database is similar. Instead of the DBD::Mysql interface

the DBD::DB2 interface has to be defined. If DB2 constants are used (for
example, SQL_MODE_READ_ONLY in the connection attributes, and others)
this has to be announced to the Perl interpreter by including “use
DBD::DB2::Constants” in the Perl program (see Example 8-22).

Example 8-22 DB2 database connection with DBI interface

use DBI;
use DBI;
use DBD::DB2;
use DBD::DB2::Constants;
my $host="localhost";
my $port="50005";
my $database="invent";
my $user="db2inst1";
my $password="password";

my $dns = "dbi:DB2:DATABASE=$database; HOSTNAME=$host; PORT=$port;

PROTOCOL=TCPIP; UID=$user; PWD=$password;";

$dbh = DBI->connect("$dns","$user","$password");

As DB2 is more powerful than MySQL, the DB2 connect statement may require a
fourth argument \%attr which contains the connection attributes, as shown in
Example 8-23.

Chapter 8. Application conversion 225

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

$dbh=DBI->connect($data_source, $user,$password, \%attr);

Example 8-23 DB2 specific connection attributes

db2_access_mode SQLL_MODE_READ_ONLY or SQL_MODE_READ_WRITE
db2_clischema Character string
db2_close_behavior SQL_CC_NO_RELEASE or SQL_CC_RELEASE
db2_connect_node Integer (must be set in DBI->connect method;
it cannot be modified afterwards)
db2_set_schema Character string
db2_db2estimate Integer
db2_db2explain One of:
SQL_DB2EXPLAIN_OFF
SQL_DB2EXPLAIN_SNAPSHOT_ON SQL_DB2EXPLAIN_MODE_ON
SQL_DB2EXPLAIN_SNAPSHOT_MODE_ON
db2_info_acctstr Character string
db2_info_applname Character string
db2_info_programname Character string
db2_info_userid Character string
db2_info_wrkstnname Character string
db2_longdata_compat Boolean
db2_quiet_mode Integer
db2_sqlerrp Character string (read only)
db2_txn_isolation One of the following:
SQL_TXN_READ_UNCOMMITTED
SQL_TXN_READ_COMMITTED
SQL_TXN_REPEATABLE_READ
SQL_TXN_SERIALIZABLE
SQL_TXN_NOCOMMIT

The DBI connect statement returns a database handle when a successful

connection is established. Otherwise, the value undef is returned. All further
communication with the database server takes places through this object.

SELECT query statements

SELECT statements are handled the same for the two pluggable interfaces. The
only conversion required is modification of the SQL syntax, if necessary, within
the prepared statement. Example 8-24 shows how the 2 interfaces use the
SELECT statement.

Example 8-24 SELECT statement used with the DBI interface

$sqlStatement = "SELECT * FROM owners;";
$($sqlStatement );
$sth->execute();
@arr = $sth->fetchrow;

226 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Calling INSERT/UPDATE/DELETE statements

Generally, the same functions can be used for INSERT, UPDATE and DELETE
statements as those discussed for the SELECT statement discussed above. For
the non-SELECT statement, DBI provides a faster substitute for
DBI::prepare/DBI::execute pair with the DBI::do function:
$rows_affected = $dbh->do($sql_statement);

For further information, refer to the DBI documentation mentioned previously.

Disconnecting from a database

The DBI interface supports the disconnect function for the database handle as
shown in Example 8-25. No change is required when converting.

Example 8-25 Disconnecting DB2 database using DBI

use DBI;
use DBD::DB2::Constants;
use DBD::DB2;
my $user="db2inst1";
my $password="password";
$dbh = DBI->connect("dbi:DB2:invent", $username, $password);

$dbh->disconnect;

8.2.2 Converting MySQL PHP applications to DB2

With PHP applications, different approaches can be used to access a MySQL
database:
򐂰 mysql for PHP
The mysql extension interface provides a procedural API to read and write to
a database.
򐂰 mysqli for PHP
The mysqli extension was introduced as a more advanced extension to the
mysql extension to support features in MySQL 4.1.3 or newer. The mysqli
provides both an object-oriented and procedure interface.
򐂰 PHP Data Objects (PDO)
MySQL databases can also be access and manipulated with the PDO and the
PDO MySQL driver.

There are 3 different approaches to connect to a DB2 application:

򐂰 IBM DB2 Extension for PHP

Chapter 8. Application conversion 227

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

The ibm_db2 extension interface provides an API to read/write from/to a

database. The mysql and mysqli extensions can be easily mapped to
ibm_db2.
򐂰 PHP Data Objects (PDO)
PDO_ODBC and PDO_IBM are object-oriented extension interfaces that are
used to connect to a database. MySQL applications using PDO can be
effortlessly converted to PDO_ODBC or PDO_IBM
򐂰 Unified ODBC
Unified ODBC is traditional procedural interface and supports multiple
database servers. The Unified ODBC extension is not optimized for DB2 and
therefore not recommended for new applications.

The following sections will cover porting PHP applications from MySQL to DB2.

For more information about developing PHP applications with DB2 check out
Developing PHP Applications for IBM Data Servers, SG24-7218.

Converting from mysql and mysqli extension library to IBM

DB2 extension for PHP
Traditionally, mysql functions were used to connect to a MySQL database from
PHP. To access MySQL V4.1 and V5 database servers from PHP 5 or higher
code, mysqli functions are used. The mysqli functions have improved methods
for accessing new features in MySQL.

If you are using either the mysql or mysqli functions to access you MySQL
database from PHP, it is relatively straightforward to convert your application to
use the ibm_db2 interface. You may also decide to perform a more complex
code rewrite and use PDO instead.

Connecting to a database
Connecting to a MySQL database with the mysql extension consists of two parts:
First a connection to the MySQL server has to be established and then a
database can be chosen.

The function specified to connect the MySQL server is given in the following
declaration:
resource mysql_connect ( [string server [, string username [, string
password [, bool new_link [, int client_flags]]]]])

Another way to connect to a MySQL server is to use the mysql_pconnect()

function, which acts like the mysql_connect(),., except this command opens a
persistent connection on the database.

228 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

The server variable in the mysql_connect() function contains the host name or
the IP address of the server that is hosting the MySQL database.

Choosing the MySQL database is done by the mysql_select_db() function:

bool mysql_select_db ( string database_name [, resource link_identifier])

Example 8-26 shows the connection part of our sample application using MySQL
database.

Example 8-26 Connecting a MySQL database with mysql extension

$host="localhost";
$user="mysql";
$pwd="password";
$database="inventory";

$conn = mysql_connect($host, $user, $pwd)or die("Couldn't connect to server. "

. mysql_error() . "\n");

$dbConn = mysql_select_db($database, $conn)or die("Couldn't select database. "

. mysql_error() .
"\n");

When connecting to a MySQL database using the mysqli extension, the

connection to the database can be done with a single command
mysqli_connect().
mysqli mysqli_connect ([ string $host = ini_get("mysqli.default_host") [,
string $username = ini_get("mysqli.default_user") [, string $passwd =
ini_get("mysqli.default_pw") [, string $dbname = "" [, int $port =
ini_get("mysqli.default_port") [, string $socket =
ini_get("mysqli.default_socket") ]]]]]] )

Example 8-27 shows an example of connecting to a MySQL database using

mysqli.

Example 8-27 Connecting a MySQL database with mysqli extension

$host="localhost";
$user="mysql";
$pwd="password";
$database="inventory";

$conn = mysqli_connect($host, $user, $pwd, $database)

or die("Couldn't connect to server. " . mysqli_connection_error() . "\n");

Just like the mysqli connections, when connecting to a DB2 database with
ibm_db2 connections to databases are done using a single db2_connect()
command.

Chapter 8. Application conversion 229

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

resource db2_connect ( string $database , string $username , string

$password [, array $options ] )

Example 8-28 shows the converted connection part of the sample application. In
our connection script, the command after the connection statement sets the
current schema which is used when querying the database. The background for
this approach is that compared to DB2, MySQL does not have schemas, nor
instances.

In MySQL, a table can be referenced by database.tablename. When no database

name is provided in the table name this defaults to the database currently in use.
In DB2, table name is defined as schema.tablename. When referencing a DB2
table, when no schema is provided, the default schema associated with the user
ID that is connected to the database is used. In an application, the set schema
statement can be used to provide a global schema name for the tables which do
not have a full table name specified. To simplify application conversion, when
porting the database, create the tables under the admin schema. By using the
set schema after database connection, we do not need to change every table
name in the application.

Example 8-28 Connect to a DB2 database

$database = 'invent';
$user = 'db2inst1';
$password = 'password';

$conn = db2_connect($database, $user, $password) or die("Couldn't connect to

server. " . db2_conn_errormsg() . "\n");

$sql="SET CURRENT SCHEMA='ADMIN'";

$sqlResult = db2_exec($conn, $sql) or die ("Couldn't execute query.... " . $sql
. "\n" );

Note: The DSN is the database name which is registered in the DB2 catalog.
Because this database is cataloged, no server information has to be declared
in the connect statement.

The persistent connection functions mysql_pconnect() and mysqli_pconnect()

can be mapped to the db2_pconnect() function.

SELECT query statements

Example 8-29 on page 230 shows querying a table in MySQL using the PHP
mysql extension functions.
Example 8-29 MySQL select example using mysql extension
$sql = "SELECT * from owners where id = $id";

230 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

$sqlResults = mysql_query($sql)
or die ("Couldn't execute query. " . $sql . "\n");
$result = mysql_fetch_array($sqlResults);
$username = $result[7];

Example 8-30 shows an example of querying a MySQL database using the

mysqli extension.
Example 8-30 MySQL select example using mysqli
$sql = "SELECT * from owners where id = $id";
$sqlResult = mysqli_query($conn, $sql)
or die ("Couldn't execute query.... " . $sql . "\n");
$result = mysqli_fetch_array($sqlResult);
$username = $result[7];

Example 8-31 shows the correspondent ibm_db2 DB2 query.

Example 8-31 DB2 SELECT example

$sql = "SELECT * from owners where id = $id";
$sqlResults = db2_exec($conn, $sql)
or die ("Couldn't execute query.... " . $sql . "\n");
$result = db2_fetch_array($sqlResults);
$username =
$result[7];

You can map the mysqli_query() function directly to the db2_exec() function, and
the db2_exec() function nearly corresponds to the mysql_query() function. The
only difference between the two functions is that db2_exec() requires two
parameters: one of which specifies the connection id returned by the
db2_connect() statement and the other is the SQL statement:
resource db2_exec ( resource $connection , string $statement [, array
$options ] )

To get each row in an array equivalent to the mysql_fetch_array() or the

mysqli_fetch_array() functions, the db2_fetch_array() function can be used
without extensive modifications.

Calling INSERT/UPDATE/DELETE statements

All three SQL commands, INSERT, UPDATE, and DELETE the same MySQL
and ibm_db2 functions are used as discussed for the SELECT statement:
mysql_query($sql)
db2_exec($db,$sql)

In Example 8-32, Example 8-33, and Example 8-34, the difference between the
MySQL and ibm_db2 functions for the INSERT statement is shown.

Chapter 8. Application conversion 231

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

Example 8-32 MySQL insert statement using mysql

$updateCMD = "UPDATE services SET serviceOwner = " . $servOwner . " WHERE ID =
" . $srvID ;
$updateOutput = mysql_query($updateCMD) ;
if($updateOutput){
$textOutput = "Service Ticket updated. \n";
}else{
$textOutput = "Service Ticket updated Failed. \n";
}
echo $textOutput . "\n";

Example 8-33 MySQL insert statement using mysqli

$updateCMD = "UPDATE services SET serviceOwner = " . $servOwner . " WHERE ID =
" . $srvID ;
$updateOutput = mysqli_query($conn, $updateCMD)
if($updateOutput){
$textOutput = "Service Ticket updated. \n";
}else{
$textOutput = "Service Ticket updated Failed. \n";
}
echo $textOutput . "\n";

Example 8-34 DB2 instert statement

$updateCMD = "UPDATE services SET serviceOwner = " . $servOwner . " WHERE ID =
" . $srvID ;
$updateOutput = db2_exec($conn, $updateCMD);
if($updateOutput){
$textOutput = "Service Ticket updated. \n";
}else{
$textOutput = "Service Ticket updated Failed. \n";
}
echo $textOutput . "\n";

These functions differ in their return values. mysql_query() and mysqli_query()

return a resource identifier only for SELECT,SHOW,EXPLAIN or DESCRIBE
statements, and return a false if the query was not executed correctly. For other
type of SQL statements, mysql_query()and mysqli_query()return true on
success and false on error. db2_exec () returns a result identifier if the SQL
command was executed successfully and false if an error occurs.

Disconnecting from a database

Example 8-35, Example 8-36, and Example 8-37 show the disconnect functions
using the mysql, mysqli and ibm_db2 extensions.

232 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Example 8-35 Disconnecting from a MySQL database using mysqli

$conn = mysql_connect($host, $user, $pwd)
or die("Couldn't connect to server. " . mysql_error() . "\n");

mysql_close($conn);

Example 8-36 Disconnecting from a MySQL database using mysqli

$conn = mysqli_connect($host, $user, $pwd, $database)
or die("Couldn't connect to server. " . mysqli_connection_error() . "\n");

mysqli_close($conn);

Example 8-37 Disconnecting form a DB2 database

$conn = db2_connect($database, $user, $password)
or die("Couldn't connect to server. " . db2_conn_errormsg() . "\n");

db2_close($conn);

The three functions above, execute the same task of disconnecting from the
database and both return true on success and false on failure. In rare cases the
return value of the mysql_close() function is not used at all (see note above),
therefore conversions can mostly be done by simply replacing the function or
inserting the new function at the end of program execution.

mysql, mysqli, and ibm_db2 function mapping

Table 8-3 maps common functions from mysql, mysqli to ibm_db2. Note that the
mysql functions only offer a single interface for retrieving database errors. The
mysqli and ibm_db2 functions provide interfaces for handling both connections
and statement errors.

Table 8-3 mysql, mysqli and ibm_db2 function mapping

mysql mysqli Ibm_db2

mysql_close mysqli_close db2_close

mysql_connect mysqli_connect db2_connect

mysql_pconnect mysqli_pconnect db2_pconnect

db2_pclose

mysql_query mysqli_query db2_exec

mysqli_stmt_prepare db2_prepare

Mysqli_stmt_execute db2_execute

mysql_fetch_array mysqli_fetch_array db2_fetch_array

mysql_fetch_assoc mysqli_fetch_assoc db2_fetch_assoc

Chapter 8. Application conversion 233

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

mysql mysqli Ibm_db2

mysql_fetch_row mysqli_fetch_row db2_fetch_row

mysql_fetch_object mysqli_fetch_object db2_fetch_object

mysql_field_name db2_field_name

mysql_errno mysqli_connection_errno db2_conn_error

mysql_error mysqli_connection_error db2_conn_errormsg

mysql_error mysqli_errno db2_stmt_error

mysql_error mysqli_error db2_stmt_errormsg

mysql_affected_rows mysqli_num_rows db2_num_rows

Converting from native MySQL library to PDO

If you are currently using PDO with MySQL, it is simple to port your application to
DB2. PDO provides a common API that you can use regardless of the database
vendor the PHP application is trying to connect to. In most cases to convert
databases all you need is change the database driver and the connection string
at which the connection to the database is established.
$db = new PDO("odbc:DSN=$database;UID=$user;PWD=$password;");
$db = new PDO("mysql:host=$hostname;dbname=$database", $user, $pass);

Connecting to a database
The syntax for connecting to a database using PDO is as follows:
$conn = new PDO( string $dsn [, string $username [, string $password [,
array $driver_options ]]] )

Example 8-38 and Example 8-39 show the difference between the syntax to
connect to a MySQL and DB2 database using PDO.

Note: The DSN is the database name, which is registered in the DB2 catalog.
Since this database is cataloged, no server information has to be declared in
the connect statement.

Example 8-38 Connecting to a MySQL database

$database = 'invent';
$user = 'db2inst1';
$password = 'password';

$conn = new PDO("mysql:host=$hostname;dbname=$database", $user, $password)

or die("Could not connect " . errorCode());

234 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Example 8-39 Connecting to a DB2 database

$database = 'invent';
$user = 'db2inst1';
$password = 'password';

$conn = new PDO("odbc:$database", $user, $password)

or die("Could not connect " . errorCode());

$query="SET CURRENT SCHEMA='ADMIN'";

$output = $conn->exec($query);

SELECT query, INSERT, UPDATE, and DELETE statements

The syntax for executing queries is exactly the same for both data servers no
change is required. Example 8-40 illustrates how to execute a query using PDO.

Example 8-40 DB2 SELECT example

$sql = "SELECT * from owners where id = '$id'";
$query = $conn->query($sql)
or die ("Couldn't execute query.... " . $sql . "\n");
foreach ($query as $query2){
$username = $query[7];
}

Disconnecting from a database

Example 8-41 shows the disconnect function using the MySQL and IBM PDO
API.

Example 8-41 Disconnecting DB2 database

$conn = new PDO("odbc:$database", $user, $password)
or die("Could not connect " . errorCode());

Sconn = null;

Converting from native MySQL library to Unified ODBC

A third extension, Unified ODBC, has historically offered access to DB2
database systems. It is not recommended that you write new applications with
this extension because ibm_db2 and pdo_ibm both offer significant performance
and stability benefits over Unified ODBC. The ibm_db2 extension API makes
porting an application previously written for Unified ODBC almost as easy as
globally changing the odbc_ function name to db2_ throughout the source code
of your application.

Chapter 8. Application conversion 235

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

Using the Unified ODBC support in PHP applications does not require a special
load of library files since support has been integrated during the compilation
process of PHP. A complete overview about the MySQL and the Unified ODBC
functions can be found in the PHP manual, available at:
http://www.php.net/docs.php

When we talk about ODBC in this section (always Unified ODBC), we refer to the
native DB2 driver. Wide similarities between syntax of Unified ODBC and other
ODBC types, and performance advantages when using Unified ODBC support,
is the reason the application conversion is discussed with the Unified ODBC
support.

Connecting to a database
When connecting to a DB2 database with ODBC, the connection is done in a
single ODBC command (odbc_connect()).
resource odbc_connect ( string dsn, string user, string password [, int
cursor_type])

Example 8-42 shows how to connect to our sample application database using
ODBC command.

Note: The DSN is the database name, which is registered in the DB2 catalog.
Since this database is cataloged, no server information has to be declared in
the connect statement.

Example 8-42 Connecting to a DB2 database

$database = 'invent';
$user = 'db2inst1';
$password = 'password';

$conn = odbc_connect($database,$user,$password)
or die("Could not connect ". odbc_errormsg());

$query="SET CURRENT SCHEMA='ADMIN'";

odbc_exec($conn,$query) or die(odbc_errormsg($db));

A complement function to the persistent connect function mysql_pconnect()is the

odbc_pconnect().

SELECT query statements

The ODBC DB2 query is shown in Example 8-43 on page 237. Refer to
Example 8-29 on page 230 and Example 8-30 on page 231 for the MySQL
SELECT statements.

236 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Example 8-43 DB2 query statement

$sql = "SELECT * FROM owners WHERE id = '$id'";
$sqlResults = odbc_exec($conn, $sql)
or die ("Couldn't execute query.... " . $sql . "\n");
odbc_fetch_into($sqlResults, $result);
$username = $result[7];

The odbc_exec() function is almost identical to the mysql_query() function. The

only difference is that the odbc_exec() function needs two parameters, one of
which is the connection id returned by the odbc_connect statement:
resource odbc_exec ( resource connection_id, string query_string)

To get each row in an array equivalent to the mysql_fetch_row() function the

odbc_fetch_into() function can be used without great modifications. There are
only syntax differences in both statements in our case.

INSERT, UPDATE, and DELETE statements

Much like MySQL, the ODBC interface can execute all three types SQL
commands, INSERT, UPDATE, and DELETE using the same function as that of
the SELECT statement:
mysql_query($sql_statement) or mysql_queryi($$db, sql_statement)
odbc_exec($db, $sql_statement)

Example 8-44 illustrates the ODBC functions for the INSERT statement. Refer to
Example 8-32 on page 232 and Example 8-33 on page 232 for the MySQL
INSERT statements.

Example 8-44 DB2 insert statement

$updateCMD = "UPDATE services SET serviceOwner = " . $servOwner . " WHERE ID =
" . $srvID ;
$updateOutput = odbc_exec($conn, $updateCMD);

if($updateOutput){
$textOutput = "Service Ticket updated. \n";
}else{
$textOutput = "Service Ticket updated Failed. \n";
}
echo $textOutput . "\n";

There is a minor difference between these functions in terms of return values:

򐂰 mysql_query() returns a resource identifier for SELECT, SHOW, EXPLAIN or
DESCRIBE statements, and false if the query is not executed correctly. For
any other SQL statements a true is returned on success and false on error.

Chapter 8. Application conversion 237

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 odbc_exec() returns a result identifier if the SQL command was executed

successfully and false if an error occurs.

Disconnecting from a database

To disconnect from a database, a difference in the return values exists between
the mysql_close() and odbc_close() statements. The function mysql_close()
returns true on success and false on failure. Whereas, the function odbc_close()
does not return a value in either case.

Example 8-45 shows the disconnect function using the ODBC library. Refer to
Example 8-28 for the MySQL disconnect function.

Example 8-45 Disconnecting from a DB2 database

$conn = odbc_connect($database,$user,$password)
or die("Could not connect ". odbc_errormsg());

odbc_close($conn);

8.2.3 Converting MySQL Ruby on Rails applications to DB2

Ruby is a fully-integrated object-oriented programming language used to develop
applications. Ruby on Rails (RoR), also known as Rails, is a framework for
developing web applications that access databases according to the
Model-View-Control architectural framework.

MySQL has two approached to connect to the to a MySQL database from Ruby:
򐂰 MySQL/Ruby API
This API is a MySQL API.
򐂰 Ruby/MySQL API
This API is developed by Ruby.

The two APIs provide the same functions the MySQL C API.

Collectively known as the IBM_DB gem, the IBM_DB Ruby driver and Rails
adapter allows Ruby applications to access the following database management
systems:
򐂰 DB2 Version 9 for Linux, UNIX, and Windows
򐂰 DB2 Universal Database (DB2 UDB) Version 8 for Linux, UNIX, and Windows
򐂰 DB2 UDB Version 5, Release 1 (and later) for AS/400® and iSeries, through
DB2 Connect
򐂰 DB2 for z/OS, Version 8 and Version 9, through DB2 Connect

238 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

򐂰 Informix Dynamic Server, Version 11.10 and later

Converting DBI interface supported code

You can easily connect to a DB2 database from Ruby on Rails by setting up the
IBM_DB Ruby driver and Rails adapter. Once the Ruby on Rails environment
has been configured, you can install the IBM_DB gem by:
򐂰 On Linux, UNIX, and Mac OS X platforms, set environment variables and
optionally source the DB2 profile
򐂰 On all supported platforms, issue the following gem command to install the
IBM_DB adapter and driver:
db2server:~ # gem install ibm_db

For further information, visit:

http://www.alphaworks.ibm.com/tech/db2onrails

Connecting to a database
Connecting to a MySQL database with the MySQ/Ruby API consists of two parts:
First a connection to the MySQL server has to be established using the connect()
function and then a database can be chosen using the select_db() function.

Example 8-46 illustrates the functions used to connect to a MySQL database

from Ruby.

Example 8-46 Connecting a MySQL database from Ruby

require 'mysql'

conn = Mysql.init()
conn.connect('localhost', 'password')
conn.select_db('test')

Connection to a DB2 database can be created using a single function, the

connect() function. The syntax for the connect function is as follows.
resource IBM_DB::connect ( string database, string username, string
password [, array options] )

Example 8-47 illustrates how to connect to a DB2 database.

Example 8-47 Connecting to a DB2 database from Ruby

require 'rubygems'
require 'ibm_db'
conn = IBM_DB.connect("test", "db2inst1", "password")

Chapter 8. Application conversion 239

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

SELECT query statements

Querying a table in MySQL using the MySQL/Ruby API functions is shown in
Example 8-48.

Example 8-48 Issuing a SELECT statement on a MySQL database from Ruby

….
results = conn.query("SELECT * FROM owners;")

while row = results.fetch_row()

puts "#{row[0]} : #{row[1]} "
end

The correspondent IBM_DB Ruby query is shown in Example 8-49 on page 240.

Example 8-49 Issuing a SELECT statement on a DB2 database from Ruby

…
results = IBM_DB.exec(conn, "SELECT * FROM admin.owners")
while row = IBM_DB.fetch_array(results)
puts "#{row[0]} : #{row[1]} "
end

The MySQL query() function can be mapped directly to the DB2 exec() function.
Both functions have similar functionality. The only difference between the two
functions is that exec() requires two parameters: one of which specifies the
connection id returned by the connect() statement and the other is the SQL
statement:
resource IBM_DB::exec ( resource connection, string statement [, array
options] )

To get each row in an array equivalent to the MySQL/Ruby fetch_row() function,

the IBM_DB fetch_array() function can be used without extensive modifications.

Calling INSERT/UPDATE/DELETE statements

Generally, the same functions can be used for INSERT, UPDATE and DELETE
statements as those discussed for the SELECT statement discussed above.

Disconnecting from a database

Both MySQL and DB2 use the a close() function to disconnect from the
database. However, the difference between the syntax is that the DB2 function
requires the connection id as a parameter for the function.

Example 8-50 and Example 8-51 show the disconnect functions using the
MySQL/Ruby API and the IBM_DB API functions.

240 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Example 8-50 Disconnecting from a MySQL database from Ruby

…
conn.close()

Example 8-51 Disconnecting from a Db2 database from Ruby

…
IBM_DB.close(conn)

For more information on developing DB2 application with Ruby on Rails review
the Developing Perl, PHP, Python, and Ruby on Rails Applications Manual
available at:
http://www.ibm.com/support/docview.wss?rs=71&uid=swg27015148

8.2.4 Converting MySQL Java applications to DB2

DB2 supports the usage of Java programming at the following levels:
򐂰 DB2 server side programming
– Java stored procedures on the DB2 server
– Java user-defined functions (UDF) on the DB2 server
򐂰 Java applications
– Java-enabled Web browser accessing DB2 using JDBC
– Java standalone application using JDBC and SQLj
򐂰 J2EE™ application server
– JavaServer™ Pages (JSP™) using a JDBC connection
– Servlets using JDBC or SQLj
– Enterprise JavaBeans™ (EJB™) using JDBC or SQLj

DB2 provides the IBM Data Server Driver for JDBC and SQLJ. This driver is a
single application driver to support the most demanding Java applications. This
agile driver can be used in either type 4 or type 2 mode. This section provides an
overview of JDBC, SQLj, and the conversion of existing MySQL Java
applications to DB2.

MySQL has an optional package MySQL Connector/J, which is a type 4 JDBC

driver. The latest Version of MySQL Connector/J implements SUN JDBC 4.0 API
for relational database access.

Java database connectivity (JDBC)

JDBC is a vendor-neutral dynamic SQL interface that provides data access for
your application through standardized Java methods. JDBC drivers provide the

Chapter 8. Application conversion 241

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

mechanics to the JDBC API to allow Java applications to access databases.

Currently JDBC API is in its fourth revision. IBM DB2 supports JDBC 3 and JDBC
4.

A JDBC application can establish a connection to a data source using the JDBC
DriverManager interface. In the following sections we discuss the changes
required within code of Java application when converting from MySQL to DB2.

IBM JDBC driver for DB2

A JDBC driver acts as an interface between a Java program and a database.
DB2 9.7 includes two JDBC drivers:
򐂰 DB2 JDBC type 2 driver
The DB2 JDBC type 2 driver, also called the native-API/partly Java driver,
which allows Java applications to make JDBC calls that are translated to Java
native methods. The Java applications that use this driver must run a DB2
client, used to communicate the JDBC requests to the DB2 server. Figure 8-2
shows a call transfer for DB2 JDBC type 2 driver. this driver can be used only
by Java applications. This driver is implemented using the DB2 CLI interface
to communicate with DB2 servers.

Figure 8-2 DB2 type 2 JDBC Driver

In order to use the DB2 JDBC type 2 driver you need following properties:
drivername="COM.ibm.db2.jdbc.app.DB2Driver"
URL="java:db2:dbname"
The user ID and password are implicitly selected according to the DB2 client
setup.

Note: The DB2 JDBC Type 2 Driver for Linux, UNIX and Windows will not
be supported in future releases. You should therefore consider switching to
the IBM Data Server Driver for JDBC and SQLJ described below.

242 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

򐂰 IBM DB2 Driver for JDBC and SQLJ

The IBM DB2 Driver for JDBC and SQLJ is a single driver that includes both
Type 2 and Type 4 connectivity. This type 4/ Native-protocol all-Java driver is
implemented in Java and it uses the Distributed Relational Database
Architecture™ (DRDA®) protocol for client/server communications.
Figure 8-3 shows the JDBC Driver usage in a Java application. DB2 Universal
JDBC Driver does not require any service on the client side and it can be
found in the db2jcc.jar package.
To use the DB2 Universal JDBC driver, you need the following properties:
drivername="com.ibm.db2.jcc.DB2Driver"
URL="java:db2://servername:serverport/dbname"

Note: Servername and serverport refer to the database server.

This driver is recommended for both applets and applications connecting to a

DB2 database.

Figure 8-3 JDBC Driver

This driver is recommended for both applets and applications connecting to a

DB2 database. For more details visit:
http://www.ibm.com/software/data/db2/ad/java.html

Embedded SQL for Java

DB2 provides embedded SQL (both static and dynamic) access to Java
applications though SQLJ APIs. DB2 SQLj support allows you to create, build,
and run embedded SQL for Java applications, applets, stored procedures, and
user-defined functions (UDFs). The SQLJ API is an extension of JDBC. JDBC
can only execute SQL statements dynamically. Statically bound SQL statements
can run faster than dynamically bound statements, hence SQLJ applications
have significant performance advantages over JDBC applications.

Chapter 8. Application conversion 243

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

Conversion of JDBC applications

Since both MySQL and DB2 comply with JDBC specifications, most Java
applications require minimal code changes. However, SQL statement changes
may still be required as discussed in DDL differences in 6.2, “Data definition
language differences” on page 122.

In this section we provide you with information on Java program conversion from
MySQL to DB2.

Loading a JDBC Driver

The first step in a Java program is to load the appropriate JDBC driver. This is
done by calling Class.forName(drivername) within your Java program.
Appropriate values for the driver names are already discussed in “IBM JDBC
driver for DB2” on page 242. Example 8-52 shows how to load the driver for
MySQL, and Example 8-53 shows the DB2 conversion.

Connecting to a database
In this part, the Java program tries to establish a connection to the given
database. This is done by calling the function DriverManager.getConnection with
the proper URL values as discussed within the driver description in “IBM JDBC
driver for DB2” on page 242. After this call, DriverManager selects the
appropriate driver from set of registered drivers, in order to connect to the
database. Example 8-52 and Example 8-53 show these steps for MySQL and
DB2 respectively.

Example 8-52 MySQL JDBC driver loading and connection

import java.sql.*;
public class inventSample {
public static void main(String[] args) throws Exception {
try {
//---------- Load the driver ---------//
Class.forName("com.mysql.jdbc.Driver");

//---------- Connect to database ---------//

Connection conn = DriverManager.getConnection
("jdbc:mysql://localhost/inventory", "mysql", "password");

//...
}
}

Example 8-53 DB2 JDBC driver loading and connection

import java.sql.*;

public class inventSample {

244 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

public static void main(String[] args) throws Exception {

try {
//---------- Load the driver ---------//
Class.forName("com.ibm.db2.jcc.DB2Driver");
try {

//---------- Connect to database ---------//

Connection conn = DriverManager.getConnection
("jdbc:db2:invent", "db2inst1", "password");
//...

}
}

Calling query statements

Since both DB2 and MySQL follow the SQL standard and the JDBC
specifications, the query execution code should not change much. Once a
establishing a connection, code changes may only be required for SQL
statements or return data types.

The JDBC API does not put any restrictions on the kind of SQL statements that
can be executed. Therefore, it becomes the responsibility of the application to
pass SQL statements compatible to the database. The connection obtained in
Example 8-52 and Example 8-53 can be used for one of the following three types
of statements depending upon the requirements:
򐂰 Statement - Simple single SQL statement
The statement can be created by using the createStatement method of the
Connection. Example 8-54 shows the usage of executeQuery with a change
for MySQL and DB2. It is evident that only changes to the SQL statement are
required.

Example 8-54 Query statement changes from MySQL to DB2

Statement s = conn.createStatement();

//---------- MySQL statement ---------//

s.executeQuery("SELECT id, firstName, lastName, loginName FROM owners WHERE
id = 501");

//---------- DB2 statement ---------//

s.executeQuery("SELECT id, firstName, lastName, loginName FROM admin.owners
WHERE id = 501");

ResultSet rs = s.getResultSet ();

while (rs.next ()){
int id = rs.getInt ("id");

Chapter 8. Application conversion 245

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

String firstName = rs.getString ("firstName");

String lastName = rs.getString ("lastName");
String userName = rs.getString ("loginName");
System.out.println (
"id = " + id
+ ", firstName = " + firstName
+ ", lastName = " + lastName
+ ", userName = " + userName);
}
rs.close ();
s.close ();

򐂰 PreparedStatement - Precompiled SQL statements

PreparedStatements are used to effectively execute particular statements
multiple times. Example 8-55 shows the conversion of the prepared
statements. Setting appropriate values of host variables is vital when using
prepared statements. In the section “Java, JDBC, and SQL data type
conversions” we discuss this further.
򐂰 callableStatement - Statement to call stored procedures.

Calling INSERT/UPDATE/DELETE statements

Any statement that updates the database or inserts and deletes a value within
the database can be executed using the executeUpdate or execute method.
Example 8-55 shows the update of a record using the prepared statement for
MySQL and its form of using the same call for DB2. As shown, these changes
are due to the MySQL databases that have been merged into one DB2 database
and grouped by the table schema. If you use DB2 SET SCHEMA = <schema_name>
for a connection, then nothing needs to be changed.

Example 8-55 MySQL Prepared statement and execute Update

//---------- MySQL prepared statement ---------//
//PreparedStatement smt = conn.prepareStatement
("UPDATE services SET serviceOwner = ? WHERE ID = ? ;");

//---------- DB2 prepared statement ---------//

PreparedStatement smt = conn.prepareStatement
("UPDATE admin.services SET serviceOwner = ? WHERE ID = ? ");

smt.setInt (1, 608);

smt.setInt (2, 108);
int count = smt.executeUpdate();
System.out.println (count + " rows were inserted");

conn.close();

246 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Java, JDBC, and SQL data type conversions

In this section we discuss how MySQL and DB2 handle column types for Java
data type conversions. As we saw in 6.1, “Data type mapping” on page 116,
MySQL and DB2 have different data types. While converting your Java
application, it is recommended to check the Java data types used to fetch your
data using JDBC. The JDBC driver converts the data exchanged between the
application and the database using the specified schema mapping, defined by
both DB2 and MySQL for their data types.

Since MySQL does not enforce strict type conversions, the Java programmer
has to take care of data loss because of round off, overflow, or precision loss.
For more detials on how MySQL is mapped to Java data types refer to:
http://dev.mysql.com/doc/refman/5.1/en/connector-j-reference-type-conversions.h
tml

On the other hand, DB2 sticks to JDBC specification, and provides a default and
recommended data type mapping as shown in Table 8-4.

Table 8-4 DB2 data types mapping to Java types

DB2 data type Java type Data type description

SMALLINT Short 16-bit signed integer

INTEGER Int 32-bit signed integer

BIGINT Long 64-bit signed integer

REAL Float Single precision floating

point

DOUBLE Double Double precision floating

point
DECIMAL java.math.BigDecimal Packed decimal

CHAR java.lang.String Fixed-length character

string of length n where n is
from 1 to 254

CHAR FOR BIT DATA byte[] Fixed-length character

string of length n where n is
from 1 to 254

VARCHAR java.lang.String Variable-length character

String

VARCHAR FOR BIT DATA byte[] Variable-length character

String

LONG VARCHAR java.lang.String Long variable-length

character string

LONG VARCHAR FOR byte[] Long variable-length

BIT DATA character string

Chapter 8. Application conversion 247

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

BLOB java.sql.Blob Large object

variable-length binary
string

CLOB java.sql.Clob Large object

variable-length character
string

DBCLOB(n) java.sql.Clob Large object

variable-length
double-byte character
string

GRAPHIC java.lang.String Fixed length double-byte

character string

VARGRAPHIC java.lang.String Non-null-terminating

varying double-byte
character string with 2 byte
string length indicator

LONG VARGRAPHIC java.lang.String Non-null-terminating

varying double-byte
character string with 2 byte
string length indicator

DB2 data type Java type Data type description

DATE java.sql.Date 10-byte character string

TIME java.sql.Time 8-byte character string

TIMESTAMP java.sql.Timestamp 26-byte character string

8.2.5 Converting MySQL C/C++ applications to DB2

DB2 provides the following programming interfaces for developing applications in
C/C++:
򐂰 Embedded SQL
򐂰 DB2 Call Level Interface (CLI)

Apart from this, DB2 uses C/C++ for server side programming for creating:
򐂰 Stored procedures on DB2 server
򐂰 User-defined functions (UDF) on DB2 server.

DB2 provides precompilers for C, C++, COBOL, Fortran, and REXX to support
embedded SQL applications. Embedded SQL applications support both static
and dynamic SQL statements. Static SQL statements require information about
all SQL statements, tables, and data types used at compile time. The application
needs to be precompiled, bound, and compiled prior to execution. In contrast,
dynamic SQL statements can be built and executed at runtime. For further

248 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

details on embedded SQL refer to Getting Started with Database Application

Development, GI11-9410-00. available at:
http://www.ibm.com/support/docview.wss?rs=71&uid=swg27015148

Another interface provided by DB2 is DB2 Call Level Interface (CLI). It is a

standard based API following the Microsoft Open Database Connectivity (ODBC)
specification and the ISO SQL/CLI standard. Both embedded SQL and DB2 CLI
support database administration and database manipulation from C/C++
applications.

MySQL provides a client library for accessing a MySQL database from C

applications. The client library is shipped with the MySQL server or can be
downloaded on its own using the MySQL connector/C. It provides features for:
򐂰 Connection mechanism
򐂰 Creation and execution of SQL queries
򐂰 Status and error reporting

The MySQL Connector/C++ (also commonly known as MySQL ++) is an

additional library for accessing MySQL databases from C++ applications.

Converting applications
MySQL C API and DB2 CLI are similar in functionality and mechanisms to
access databases. Both use the function call to pass dynamic SQL statements
and do not need to be precompiled. We recommend converting MySQL C
applications to DB2 CLI. This section describes conversion changes for various
levels of the application:

Connecting to the server

The first step in converting MySQL C applications is to change the include
information, initialize variables, and replace the MySQL connection with a DB2
connection. Example 8-56 shows a typical MySQL C program to initiate MySQL
variables, create a connection, and terminate the connection.

Example 8-56 MySQL C application, initialize MySQL and create connection

#include <mysql/mysql.h> /* Include MySQL variable and function definition */
#include <stdio.h>

int main(){

MYSQL *conn; /* pointer to connection handler */

connection = mysql_init(NULL);

if(!mysql_real_connect (
conn, /* pointer to connection handler */
NULL, /* host to connect, default localhost*/
"mysql", /* user name, default local user*/

Chapter 8. Application conversion 249

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

"password", /* password, default none*/

"inventory", /* database name*/
0, /* port */
NULL, /* socket */
0 /* flags*/
) )
{
printf("Connection Failed \n");
fprintf(stderr, "%s\n", mysql_error(conn));
} else{
printf("Successful connection to database.\n");
}
}
mysql_close(conn);
}

Figure 8-4 on page 250 shows a similar task using DB2 CLI. It shows the
initialization tasks, which consist of: allocation and initializing the environment
and connection handlers; creating the connection; processing of transactions;
and terminating the connection and removing of handlers.

Figure 8-4 DB2 CLI activities

Example 8-57 shows the implementation of the task defined by the figure above.

250 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Example 8-57 DB2 CLI application, connecting to a database

#include <sqlcli.h> /* Include DB2 CLI variable and function definition */ int main()
{
SQLRETURN ret = SQL_SUCCESS; int rc = 0;
SQLHANDLE henv; /* environment handle */
SQLHANDLE hdbc; /* connection handle */

/* Allocate an environment handle */

ret = SQLAllocHandle(SQL_HANDLE_ENV, SQL_NULL_HANDLE, &henv);
if (ret != SQL_SUCCESS){
/* handle error */
return 1;
}

/* Allocate a connection handle */

ret = SQLAllocHandle(SQL_HANDLE_DBC, henv, &hdbc);

if (ret != SQL_SUCCESS){
/* handle error */
return 1;
}

/* connect to the database */

ret = SQLConnect(hdbc,
(SQLCHAR *)"invent", SQL_NTS,
(SQLCHAR *)"db2inst1" /*user*/, SQL_NTS,
(SQLCHAR *)"password" /* password*/, SQL_NTS);
if (ret != SQL_SUCCESS){
/* handle error */
return 1;
}

/* disconnect from the database */

ret = SQLDisconnect(hdbc);
if (ret != SQL_SUCCESS){
/* handle error */
return 1;
}

/* free connection handle */

ret = SQLFreeHandle(SQL_HANDLE_DBC, hdbc);
if (ret != SQL_SUCCESS){
/* handle error */
return 1;
}

/* free environment handle */

ret = SQLFreeHandle(SQL_HANDLE_ENV, henv);
if (ret != SQL_SUCCESS) {
/* handle error */
return 1;
}

Chapter 8. Application conversion 251

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

exit(0);
}

Processing a query
A typical MySQL C API program involves three steps in query processing:
򐂰 Query construction
Depending upon your requirement you can construct a null terminated string
or counted length string for the query:
char *query;
򐂰 Execution of the query
For executing the query you can use mysql_real_query() for a counted length
query string or the mysql_query() for a null terminated query string.
Example 8-58 shows the processing of a query with both mysql_real_query()
and mysql_query() method calls.
򐂰 Processing of the returned results
After executing the query, the final step is to process the results. All the
statements except SELECT, SHOW, DESCRIBE, and EXPLAIN do not return
results. For these statements, mysql provides the mysql_affected_rows()
function for accessing the number of rows affected.
If your query returns a result set, follow these steps for the result processing:
a. Generate the result set using mysql_store_result() or mysql_use_result().
b. Fetch each row using mysql_fetch_row().
c. Release the result set using mysql_free_result().

Example 8-58 shows an example of both MySQL queries - some that return a set
of results and others that do not.

Example 8-58 MySQL query processing

MYSQL_RES *result;
if (mysql_query(conn, "SELECT * FROM owners WHERE id = 501")) {
fprintf(stderr, "%s\n", mysql_error(conn));
return 1;
}
result = mysql_use_result(conn);

/* output table name */

printf("MySQL Tables in mysql database:\n");
while ((row = mysql_fetch_row(result)) != NULL){
printf("firstName: %s \n", row[1]);
printf("lastName: %s \n", row[2]);
printf("email: %s \n", row[3]);
}

252 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

mysql_free_result(result);
}
}

On the other hand DB2 CLI provides a more comprehensive set of APIs for doing
similar tasks. One of the essential parts of DB2 CLI is transaction processing,
which is supported by all the tables in DB2. Figure 8-5 on page 253 shows the
typical order of function calls of query processing.

Figure 8-5 DB2 Query processing

DB2 CLI query processing involves the following steps:

1. Allocating a statement handle

Chapter 8. Application conversion 253

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

A statement handle tracks execution of the query for a particular connection.

This handle can be allocated using SQLAllocHandle() with a HandleType of
SQL_HANDLE_STMT function.
2. Preparing and executing SQL statements
DB2 provides two ways for preparing and executing the query:
– Prepare and execute as separate steps
If you plan to execute the same query multiple times with different
parameters, you can use this technique. This involves the following steps:
Preparing a query using SQLPrepare(), binding of parameters using
SQLBindParameter(), and executing this query using SQLExecute().
Example 8-59 shows an example for prepared statements.
– Prepare and execute in a single step
If your query is executed only once, then you can use SQLExecDirect() to
directly call, prepare, and execute in a single step. Example 8-60 shows
the usage of this method.
3. Processing results
Processing query results involves binding application variables to columns of
a result set, and then fetching the rows of data into the application variables.
This is done by calling SQLBindCol() followed by SQLFetch() as shown in
Example 8-59.
Another way to get data without binding the column is by calling SQLFetch()
and SQLGetData(), this technique is used in Example 8-60.
4. Committing or rolling back
DB2 supports two commit modes: auto-commit and manual commit. This can
be set using SQLSetConnectAttr() with the parameter
SQL_AUTOCOMMIT_ON or SQL_AUTOCOMMIT_OFF. If a transaction is
set to SQL_AUTOCOMMIT_OFF it is the programmer's responsibility to end
the transaction. This can be done using SQLEndTran() to either rollback or
commit the transaction using parameter SQL_COMMIT or SQL_ROLLBACK.
5. Deallocating statement handle
This requires unbinding of variables, columns, or cursors (if allocated) using
SQLFreeStmt() with the option of SQL_CLOSE, SQL_UNBIND or
SQL_RESET_PARAMS, and then finally calling SQLFreeHandle() to
deallocate the statement handle.

Example 8-59 DB2 CLI prepared statement with column binding, auto commit on
SQLHANDLE hstmt; /* statement handle */
SQLCHAR firstName [TEXT_SIZE];
SQLCHAR lastName [TEXT_SIZE];
SQLCHAR email [TEXT_SIZE];

254 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

SQLINTEGER id = 501;

/* SQL statements to execute */

SQLCHAR *stmt = (SQLCHAR *)"SELECT firstName, lastName, email FROM admin.owners WHERE
id = ?";

/* set AUTOCOMMIT on */
ret = SQLSetConnectAttr(hdbc, SQL_ATTR_AUTOCOMMIT, (SQLPOINTER)SQL_AUTOCOMMIT_ON,
SQL_NTS);

if (ret != SQL_SUCCESS) {
/* handle error */
}

/* allocate a statement handle */

ret = SQLAllocHandle(SQL_HANDLE_STMT, hdbc, &hstmt);
if (ret != SQL_SUCCESS) {
/* handle error */
}

/* prepare the statement */

ret = SQLPrepare(hstmt, stmt, SQL_NTS);
if (ret != SQL_SUCCESS) {
/* handle error */
}
/* bind parameter1 to the statement */
ret = SQLBindParameter(hstmt, 1, SQL_PARAM_INPUT, SQL_C_UBIGINT, SQL_CHAR, 0, 0, &id,
0, NULL);
if (ret != SQL_SUCCESS){
/* handle error */
}

ret = SQLBindCol(hstmt, 1, SQL_C_CHAR, (SQLPOINTER)firstName, sizeof(firstName) + 1,

NULL);
ret = SQLBindCol(hstmt, 2, SQL_C_CHAR, (SQLPOINTER)lastName, sizeof(lastName) + 1,
NULL);
ret = SQLBindCol(hstmt, 3, SQL_C_CHAR, (SQLPOINTER)email, sizeof(email) + 1, NULL);

/* execute the statement */

ret = SQLExecute(hstmt);
if (ret != SQL_SUCCESS){
/* handle error */
}

/* fetch each row and display */

ret= SQLFetch(hstmt);

if(ret == SQL_NO_DATA_FOUND){
printf("No data found");
}
while(ret != SQL_NO_DATA_FOUND){
printf("First name: %s \n",firstName);
printf("Last name: %s \n",lastName);

Chapter 8. Application conversion 255

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

printf("email: %s \n", email);

ret=SQLFetch(hstmt);
}
ret = SQLFreeStmt(hstmt, SQL_UNBIND);
if (ret != SQL_SUCCESS) {
/* handle error */
}
/* free the statement handle */
ret = SQLFreeHandle(SQL_HANDLE_STMT, hstmt);
if (ret != SQL_SUCCESS) {
/* handle error */
return 1;
}

Example 8-60 DB2 CLI prepare/execute in one step with SQLGetData and manual commit
SQLHANDLE hstmt; /* statement handle */
SQLCHAR firstName [TEXT_SIZE];
SQLCHAR lastName [TEXT_SIZE];
SQLCHAR email [TEXT_SIZE];

/* SQL statements to execute */

SQLCHAR *stmt1 = (SQLCHAR *)"SELECT firstName, lastName, email FROM admin.owners
WHERE id = 501";

/* set AUTOCOMMIT on */
ret = SQLSetConnectAttr(hdbc,
SQL_ATTR_AUTOCOMMIT, (SQLPOINTER)SQL_AUTOCOMMIT_OFF, SQL_NTS);
if (ret != SQL_SUCCESS) {
/* handle error */
}

/* allocate a statement handle */

ret = SQLAllocHandle(SQL_HANDLE_STMT, hdbc, &hstmt);
if (ret != SQL_SUCCESS) {
/* handle error */
}

/* execute statement 1 directly */

ret = SQLExecDirect(hstmt, stmt1, SQL_NTS);
if (ret != SQL_SUCCESS) {

if(ret == SQL_NO_DATA_FOUND) {
printf("No data found");
}
int count = 0;
while(ret != SQL_NO_DATA_FOUND) {

/* get data from column 1 */

ret = SQLGetData(hstmt, 1, SQL_C_CHAR, (SQLPOINTER)firstName, sizeof(firstName) + 1,
NULL);
count = count + 1;

256 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

printf("COUNT %i \n", count);

printf("First Name: %s \n", firstName);
if (ret != SQL_SUCCESS) {
/* handle error */
}

/* get data from column 2 */

ret = SQLGetData(hstmt, 2, SQL_C_CHAR, (SQLPOINTER)lastName, sizeof(lastName) + 1,
NULL);
printf("Last Name: %s \n", lastName);
if (ret != SQL_SUCCESS) {
/* handle error */
}

/* get data from column 3 */

ret = SQLGetData(hstmt, 3, SQL_C_CHAR, (SQLPOINTER)email, sizeof(email) + 1, NULL);
printf("email : %s \n", email);
if (ret != SQL_SUCCESS) {
/* handle error */
}

ret=SQLFetch(hstmt);
}
ret = SQLEndTran( SQL_HANDLE_DBC, hdbc, SQL_ROLLBACK );
if (ret != SQL_SUCCESS) {
/* handle error */
return 1;
}

/* free the statement handle */

ret = SQLFreeHandle(SQL_HANDLE_STMT, hstmt);
if (ret != SQL_SUCCESS){
/* handle error */
return 1;
}

For all applications using the MySQL Connector/C++, you may want to consider
converting these to DB2 CLI. The typical conversion process would remain the
same as both MySQL C and MySQL C++ use the same flow of the program.

8.2.6 Converting Connector/ODBC applications to DB2

MySQL supports the ODBC database API to connect to a MySQL database
server using the optional product called MySQL Connector/ODBC which is also
formally known as MyODBC. Support for Connector/ODBC is available at two
levels:
򐂰 Connector/ODBC 3.51 is a 32-bit ODBC Driver supporting ODBC 3.51
specification

Chapter 8. Application conversion 257

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 Connector/ODBC 5.1 which supports 64-bit platforms and improves on

Connector/ODBC 3.51 support.

Since DB2 CLI is also based on the ODBC specification, and you can build
ODBC applications without using any ODBC driver manager, so the application
conversion is quite easy. All you must do is use DB2's ODBC driver to link to
your application with libdb2. The DB2 CLI driver also acts as an ODBC driver
when loaded by an ODBC driver manager. DB2 CLI conforms to ODBC 3.51.

Figure 8-6 on page 258 shows the MySQL driver and DB2 ODBC driver in the
ODBC scenario. This shows the simplicity of converting an application using an
ODBC driver to another driver. It also shows various components involved in the
ODBC application and how they are mapped from MySQL Connector/ODBC to
DB2 ODBC.

Figure 8-6 ODBC application conversion from MyODBC to DB2 ODBC Driver

Typically, the MySQL Connector/ODBC connector has five components:

򐂰 Application
Since both MyODBC and DB2 CLI are based on ODBC specification,
applications do not require many changes. Though you have to perform some
tasks such as:
– Changing SQL queries
– Changes in transaction management
– Proprietary method changes, deviating from ODBC or other methods.

258 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

򐂰 Driver manager
IBM Data Server CLI and ODBC Driver do not come with ODBC Driver
Manager. When using an ODBC application you must ensure that an ODBC
Driver Manager is installed and users using ODBC have access to it. The
following ODBC Driver Managers can be configured to work with the IBM
Data Server CLI and ODBC Driver:
– unixODBC driver manager
The unixODBC Driver Manager is an open source ODBC driver manager
supported for DB2 ODBC applications on all supported Linux and UNIX
operating systems.
– Microsoft ODBC driver manager
The Microsoft ODBC driver manager can be used for connections to
remote DB2 databases when using TCP/IP network.
– DataDirect ODBC driver manager
The DataDirect ODBC driver manager for DB2 can be used for
connections to the DB2 database.
򐂰 Connector/ODBC
As shown in Figure 8-6 on page 258, you are not required to use
Connector/ODBC now, instead you use the DB2 ODBC Driver.
򐂰 ODBC Configuration
The ODBC Driver Manager uses two initialization files:
– /etc/unixODBC/odbcinst.ini, in which you must add the following:
[IBM DB2 ODBC DRIVER]
Driver=/home/<instance name>/sqllib/lib/db2.o
– /home/<instance name>/.odbc.ini, in which you must configure the data
source; for setting up a data source you must add the following:
In [ODBC Data Source] stanza add
invent= IBM DB2 ODBC DRIVER
Add [invent] stanza with
Driver=/home/<instance name>/sqllib/lib/libdb2.so
Description= invent DB2 ODBC Database
򐂰 MySQL Server
The MySQL database server is replaced by the DB2 server; this is discussed
in detail in previous chapters.

You may optionally configure DB2 ODBC Driver to modify the behavior of the
DB2 ODBC Driver. This can be done by changing the db2cli.ini file.

Chapter 8. Application conversion 259

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

8.2.7 Condition handling in DB2

Condition handling is one of the least sparkling topics in the application
development area. Foreseeing and handling exceptions in the application is
paramount for a robust, industrial strength programming style. This section
attempts to give an introduction to condition handing in DB2.

DB2 Error checking with SQLCODE and SQLSTATE

Each time an SQL statement is executed two values are returned by the DB2
engine and placed in the SQL communication area (SQLCA): SQLCODE and
SQLSTATE.
Within your application program you can retrieve these values to determine the
state of the previously executed SQL statement. These identifiers provide more
detailed information about the condition of the statement.

SQLCODE is the variable conventionally used for error handling in applications

coded against the DB2 family. Therefore, SQLCODE is probably of the finest
granularity when it comes to DB2 exception handling.

However, the value for SQLCODE is IBM defined. To achieve the highest
portability of applications, you should only build dependencies on a subset of
DB2 SQLSTATEs that are defined by ODBC Version 3 and ISO SQL/CLI
specifications. Whenever you build your exception handling on IBM supplied
SQLSTATEs or SQLCODEs, the dependencies should be carefully documented.
The specifications can be found using the search words ISO/IEC and standards
9075-1, 9075-2, and 9075-3 for SQL Foundation.

SQLSTATE is a five character string conforming to the ANSI SQL92 standard.

The first two characters are known as the SQLSTATE class code. For example:
򐂰 00 means successful completion
򐂰 01 is a warning
򐂰 HY is generated by the DB2 CLI (call level interface) or ODBC driver
򐂰 IM is generated by the ODBC driver manager

For example, if your application would signal SQLSTATE 23000, the DB2
description reports an integrity constraint violation, which is similar to MySQL's
rudimentary description ER NON UNIQ ERROR or ER DUP KEY. Hence,
condition handling for both database management systems could almost
execute the same code.

SQLSTATE for DB2 CLI

Follow these guidelines for using SQLSTATEs within your CLI application:

260 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

򐂰 Always check the function return code before calling SQLGetDiagRec() to

determine if diagnostic information is available. Refer to the IBM DB2 Call
Level Interface Guide and Reference, Volume 1, SC10-4224-00 for more
information on this API.
򐂰 Use the SQLSTATEs rather than the native error code.
򐂰 To increase your application's portability, only build dependencies on the
subset of DB2 CLI SQLSTATEs that are defined by the ODBC version 3 and
ISO SQL/CLI specifications, and return the additional ones as information
only.
It may be useful to build dependencies on the class (the first two characters)
of the SQLSTATEs.
򐂰 For maximum diagnostic information, return the text message along with the
SQLSTATE (if applicable, the text message will also include the IBM defined
SQLSTATE). It is also useful for the application to print out the name of the
function that returns the error.
򐂰 Ensure that the string allocated for the SQLSTATE includes space for the null
termination character returned by DB2 CLI.

The code segment from utilcli.c in Example 8-61shows how diagnostic

information such as SQLSTATEs can be retrieved and displayed.

Example 8-61 Handling SQLSTATE in CLI

void HandleDiagnosticsPrint(SQLSMALLINT htype, /* type identifier */
SQLHANDLE hndl /* handle */ )
{
SQLCHAR message[SQL_MAX_MESSAGE_LENGTH + 1];
SQLCHAR sqlstate[SQL_SQLSTATE_SIZE + 1];
SQLINTEGER sqlcode;
SQLSMALLINT length, i;

i = 1;

/* get multiple field settings of diagnostic record */

while (SQLGetDiagRec(htype,
hndl,
i,
sqlstate,
&sqlcode,
message,
SQL_MAX_MESSAGE_LENGTH + 1,
&length) == SQL_SUCCESS)
{
printf("\n SQLSTATE= %s\n", sqlstate);
printf("Native Error Code = %ld\n", sqlcode);
printf("%s\n", message);
i++;
}

Chapter 8. Application conversion 261

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

printf("-------------------------\n");
}

Note: The code snippets provided in this chapter are for illustration purposes
only. utilcli.c is sample code shipped with DB2 and can be found in the
SQLLIB/samples directory.

Handling SQL errors in an SQLj application

SQLj clauses use the JDBC class java.sql.SQLException for error handling.
SQLj generates an SQLException under the following circumstances:
򐂰 When any SQL statement returns a negative SQL error code
򐂰 When a SELECT INTO SQL statement returns a +100 SQL error code

You can use the getErrorCode method to retrieve SQL error codes and the
getSQLState method to retrieve SQLSTATEs.

To handle SQL errors in your SQLj application, import the java.sql.SQLException

class, and use the Java error handling try/catch blocks to modify the program
flow when an SQL error occurs (see Example 8-62).

Example 8-62 SQL Exception with SQLj

try {
#sql [ctxt] {SELECT LASTNAME INTO :empname
FROM EMPLOYEE WHERE EMPNO='000010'};
}
catch(SQLException e) {
System.out.println("Error code returned: " + e.getErrorCode());
}

For exception handling in Java it is important to know that DB2 provides several
types of JDBC drivers with slightly different characteristics. With the DB2
Universal JDBC Driver, you can retrieve the SQLCA. For the DB2 JDBC type 2
driver for Linux, UNIX, and Windows (DB2 JDBC type 2 driver), use the standard
SQLException to retrieve SQL error information.

SQLException under the IBM Data Server Driver for JDBC and
SQLJ
As in all Java programs, error handling is done using try/catch blocks. Methods
throw exceptions when an error occurs, and the code in the catch block handles
those exceptions.

262 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

JDBC provides the SQLException class for handling errors. All JDBC methods
throw an instance of SQLException when an error occurs during their execution.
According to the JDBC specification, an SQLException object contains the
following information:
򐂰 A string object that contains a description of the error or null
򐂰 A string object that contains the SQLSTATE or null
򐂰 An integer value that contains an error code
򐂰 A pointer to the next SQLException or null

IBM DB2 Driver for JDBC and SQLJ provides an extension to the SQLException
class, which gives you more information about errors that occur when DB2 is
accessed. If the JDBC driver detects an error, the SQLException class provides
you the same information as the standard SQLException class. However, if DB2
detects the error, the SQLException class provides you the standard information,
along with the contents of the SQLCA that DB2 returns. If you plan to run your
JDBC applications only on a system that uses the IBM DB2 Driver for JDBC and
SQLJ, you can use this extended SQLException class.

Under the IBM DB2 Driver for JDBC and SQLJ, SQLExceptions from DB2
implement the com.ibm.db2.jcc.DB2Diagnosable interface. An SQLException
from DB2 contains the following information:
򐂰 A java.lang.Throwable object that caused the SQLException or null if no such
object exists. The java.lang.Throwable class is the superclass of all errors
and exceptions in the Java language.
򐂰 The information that is provided by a standard SQLException
򐂰 An object of DB2-defined type DB2Sqlca that contains the SQLCA. This
object contains the following objects:
– An INT value that contains an SQL error code
– A String object that contains the SQLERRMC values
– A String object that contains the SQLERRP value
– An array of INT values that contains the SQLERRD values
– An array of CHAR values that contains the SQLWARN values
– A String object that contains the SQLSTATE

The basic steps for handling an SQLException in a JDBC program that runs
under the IBM DB2 Driver for JDBC and SQLJ are:
1. Import the required classes for DB2 error handling,
com.ibm.db2.jcc.DB2Diagnosable for getting diagnostic data, and
com.ibm.db2.jcc.DB2Sqlca for receiving error messages.
2. In your code catch SQLException and use it to get SQLCA. This is allowed
only if the exception thrown is an instance of the DB2Diagnosable class.

Chapter 8. Application conversion 263

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

3. Once you have DB2Sqlca, it can be used to get SQLCODE, messages, SQL
errors, and winnings as shown in Example 8-63.

Example 8-63 Processing an SQLEXception under the Universal JDBC Driver

import java.sql.*;
import com.ibm.db2.jcc.DB2Diagnosable;
import com.ibm.db2.jcc.DB2Sqlca;

……
try {
// Code that could throw SQLExceptions }
….
catch(SQLException sqle) {
while(sqle != null) {
if (sqle instanceof DB2Diagnosable) {
DB2Sqlca sqlca = ((DB2Diagnosable)sqle).getSqlca();
if (sqlca != null) {
System.err.println ("SqlCode: " + sqlca.getSqlCode());
System.err.println ("SQLERRMC: " + sqlca.getSqlErrmc());
System.err.println ("SQLERRP: " + sqlca.getSqlErrp() );
String[] sqlErrmcTokens = sqlca.getSqlErrmcTokens();
for (int i=0; i< sqlErrmcTokens.length; i++) {
System.err.println (" token " + i + ": " + sqlErrmcTokens[i]);
}
int[] sqlErrd = sqlca.getSqlErrd();
char[] sqlWarn = sqlca.getSqlWarn();
System.err.println ("SQLSTATE: " + sqlca.getSqlState());
System.err.println ("message: " + sqlca.getMessage());
}
}
sqle=sqle.getNextException();
}
}

Error handling using the WHENEVER statement

The WHENEVER statement causes the SQL precompiler to generate source
code that directs the application to go to a specified label if either an error, a
warning, or no rows are found during execution. The statement, WHENEVER,
affects all subsequent SQL statements until another WHENEVER statement
alters the situation.

The WHENEVER statement has three basic forms:

EXEC SQL WHENEVER SQLERROR action
EXEC SQL WHENEVER SQLWARNING action
EXEC SQL WHENEVER NOT FOUND action

In the above statements:

򐂰 SQLERROR: Identifies any condition where SQLCODE < 0.

264 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

򐂰 SQLWARNING: Identifies any condition where SQLWARN(0) = W or

SQLCODE > 0 but is not equal to 100.
򐂰 NOT FOUND: Identifies any condition where SQLCODE = 100.

In each case, the action includes either of the following:

򐂰 CONTINUE: Indicates to continue with the next instruction in the application.
򐂰 GO TO label: Indicates to go to the statement immediately following the label
specified after GO TO (also commonly used as GOTO.)

When the WHENEVER statement is not used, by default execution continues

even if an error, warning, or exception condition occurs.

The statement WHENEVER must be used prior to the SQL statements that will
be affected. Otherwise, the precompiler does not know that additional
error-handling code should be generated for executable SQL statements. You
can have any combination of the three basic forms active at any time. The order
in which you declare the three forms is not significant.

To avoid an infinite looping situation, ensure that you undo the WHENEVER
handling prior to execution of any SQL statements within the handler. You can do
this by using the WHENEVER SQLERROR CONTINUE statement.

Declaring the SQLCA for error handling

You can declare the SQLCA in your application program so that the database
manager can return information to your application. When you preprocess your
program, the database manager inserts host language variable declarations in
place of the INCLUDE SQLCA statement. The system communicates with your
program using the variables for warning flags, error codes, and diagnostic
information.

After executing each SQL statement, the system issues a return code in both
SQLCODE and SQLSTATE. SQLCODE is an integer value that summarizes the
execution of the statement, and SQLSTATE is a character field that provides
common error codes across IBM's relational database products. SQLSTATE
also conforms to the ISO/ANS SQL92 and FIPS 127-2 standard.

Note that if SQLCODE is less than 0, it means an error has occurred and the
statement has not been processed. If the SQLCODE is greater than 0, it means a
warning has been issued, but the statement is still processed.

For a DB2 application written in C or C++, if the application is created using

multiple source files, only one of the files should include the EXEC SQL
INCLUDE SQLCA statement to avoid multiple definitions of the SQLCA. The
remaining source files should use the following lines:

Chapter 8. Application conversion 265

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

#include "sqlca.h"
extern struct sqlca sqlca;

Condition handling in DB2 stored procedure

Similar to MySQL, DB2 defines condition handling using a stored procedure.
Although the support is similar, it seems appropriate to include a few examples of
DB2 condition handling. For detailed information on condition handlers you may
refer to SQL Procedural Languages: Application Enablement and Support,
SC09-4827.

The general form of a handler declaration is:

---DECLARE-----+-CONTINUE-+---- HANDLER-- FOR---condition------>
+-EXIT---------+
+-UNDO---------+
>-----------SQL-procedure-statement-----------------------------|

When DB2 raises a condition that matches a condition, DB2 passes control to
the condition handler. The condition handler performs the action indicated by
handler-type, and then executes SQL-procedure-statement.

DB2 provides three general conditions:

򐂰 NOT FOUND:
Identifies a condition resulting in an SQLCODE of +100 or an SQLSTATE
beginning with the characters '02'.
򐂰 SQLEXCEPTION:
Identifies any condition that results in a negative SQLCODE.
򐂰 SQLWARNING:
Identifies any condition that results in a warning condition (SQLWARN0 is
'W'), or that results in a positive SQL return code other than +100. The
corresponding SQLSTATE value will begin with the characters '01'.

You can also use the DECLARE statement to define your own condition for a
specific SQLSTATE.

Example 8-64 shows the general flow of the condition handler in a stored
procedure.

Example 8-64 General example for condition handling

Begin
declare exit handler
for sqlexception
begin
statement3;

266 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

statement4;
end;

statement1;
statement2;
End

Example 8-65 shows a CONTINUE handler for delete and update operations on
a table named EMP. Again, note that this code is solely intended for illustration
purposes.

Example 8-65 Example of a DB2 CONTINUE handler

CREATE PROCEDURE PROC1()
LANGUAGE SQL
BEGIN
DECLARE SQLCODE, v_error INT;
DECLARE CONTINUE HANDLER FOR
SQLEXCEPTION,
SET v_error = SQLCODE;

DELETE FROM emp

WHERE empno BETWEEN 100 and 200;
IF (v_error = -147 ) THEN
INSERT . . .
UPDATE staff SET salary = salary * 1.25;
IF (v_error <> 0 ) THEN
RETURN -1;
END IF;
END

8.2.8 Special conversions

MySQL provides access control on the host level. That means specific privileges
are granted depending on the host from which the user connects. DB2 does not
provide this control mechanism. Therefore, if you use this MySQL feature, you
have to implement a workaround in the application. Applications which use the
MySQL host authentication feature to control user privileges on a database or
global level will require code change.

There are many ways to implement this authentication mechanism on the

application level. Here we demonstrate a workaround using a simple example
application that has two functions SELECT and INSERT, which use the MySQL
security feature to limit selecting and inserting data on the host level.
Example 8-66 shows the MySQL host access data for four users controlled by
our example.

Chapter 8. Application conversion 267

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

Example 8-66 MySQL host access date

mysql> select user, host, select_priv, insert_priv from user;
+---------------+------------------+-------------+-------------+
| user | host | select_priv | insert_priv |
+---------------+------------------+-------------+--------------+
| root | localhost | N | N |
| | localhost | N | N |
| user01 | % _____ | Y _____ | N |
| user02 | %.ibm.com | N | Y | |
| inventAppUser | localhost | Y | Y |
| inventAppUser | %.ibm.com | Y ________ | Y |
+---------------+------------------+-------------+--------------+

This information has to be ported into a DB2 table. When a user attempts to
access the data in the DB2 database, the application will verify each user's
database access rights along with the host system information where one is
connecting from.

Two tables are needed for our DB2 conversion: one to store user privilege
information ported from MySQL and one is a working table. The table definitions
and some sample values are shown in Example 8-67 on page 268.

Example 8-67 Creation of the tables for host authentication

-- script for creating the tables used by our example application

-- connect to the database

connect to invent user db2inst1 using password;

-- table ACCESSLIST
-- it stores access rights for specific users connecting from specific hosts
-- remark: there should be different access-flags for different functions

-- fields:
-- username, whom access to the function should be granted
-- hostname or ip-address, from which the user must connect
-- select access flag (Y/N), if SELECT is granted
-- insert access flag (Y/N), if INSERT is granted

drop table ACCESSLIST;

create table ACCESSLIST (
USERNAMEvarchar(20),
HOSTvarchar(30),
ACCESS_SELchar(1),
ACCESS_INSchar(1)
);
-- insert some sample values, according to the MySQL values (see above)
insert into ACCESSLIST values('user01', '%', 'Y', 'N');
insert into ACCESSLIST values('user02', '%.ibm.com', 'N', 'Y');
insert into ACCESSLIST values('inventAppUser', 'localhost', 'Y', 'Y');

268 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

insert into ACCESSLIST values('inventAppUser', '%.ibm.com', 'Y', 'Y');

-- table APPLACCESS
-- it stores the info about users and their host asking for access
-- this table is filled automatically by the sample application

-- fields:
-- username, who asks for access to the function
-- hostname, from which the user connects
-- ip-address, from which the user connects
-- timestamp, when the user asks for access

drop table APPLACCESS;

create table APPLACCESS (
USERNAME varchar(8),
HOSTNAME varchar(30),
IPADDR varchar(30),
TS timestamp
);

The tables ACCESSLIST and APPLACCESS are used by the authentication

mechanism:
򐂰 ACCESSLIST table
This table stores all the combinations of users and hosts (either name or
IP-address) and specific database access privileges are granted to user and
host combinations. In our example, we control two access rights: SELECT
and INSERT, so that we have two access fields ACCESS_SEL and
ACCESS_INS. In a real world application more functions would probably be
controlled, so this table should be expanded to have one ACCESS field for
each function to be controlled based on the corresponding MySQL privilege.
We insert some sample values into this table for demonstration purposes.
򐂰 APPLACCESS table
This table is filled by the authentication application during runtime. When a
user asks for access, the application inserts the user ID, the host name, and
the IP-address from where the user connects. The timestamp is used as a
key in this table as records are not deleted from it.

The application code is listed in Example 8-68; remember that the code is just for
demonstration purposes.

Example 8-68 Authentication mechanism example

import java.lang.*;
import java.io.*;
import java.sql.*;
import java.net.*;
public class AccessControl

Chapter 8. Application conversion 269

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

{
// in our example we use fixed values, you should make this variable
private static final String DB2DB = "invent"; // database name
private static final String DB2USR = "db2inst1"; // database user
private static final String DB2PWD = "password"; // database password

private static Connection db2Conn; // DB2 connection object

public static void main(String[] args) throws SQLException, Exception

{
Class.forName("COM.ibm.db2.jdbc.app.DB2Driver").newInstance();

db2Conn=DriverManager.getConnection("jdbc:db2:"+DB2DB,DB2USR,DB2PWD);

// This example shows how to verify accesss to specific functions

// depending on the host where the program is executed.
// We assume that this program runs locally and connects directly to DB2
// In a client/server-app, the server must detect the client's host.

// Java provides two methods for this purpose:

// socket.getInetAddress() ... for socket-connections
// request.getRemoteHost() ... for HTTP-connections

// In this example the username is the same as the DB2 user.

// The application user and the database user could also be different.
// This example implements no error handling.

// This example shows access to two specific functions:

// SELECT...this is indicated by mode='SEL'
// INSERT...this is indicated by mode='INS'
// mode should be a variable parameter

String mode = "SEL";

String insertStr = "";

InetAddress addr = InetAddress.getLocalHost(); // appl executes locally

String ipname = addr.getHostName(); // get the hostname
String ipaddr = addr.getHostAddress(); // get the ip-address
System.out.println("Host name = " + ipname);
System.out.println("Host addr = " + ipaddr);
// get records for the specified username with access to the function
String query1 = "select HOST, current timestamp as TS from ACCESSLIST "+
"where USERNAME='" + DB2USR + "' and ACCESS_" + mode + "='Y'";
System.out.println("Query = " + query1);
PreparedStatement ps1 = db2Conn.prepareStatement(query1);
ResultSet rs = ps1.executeQuery(); // run the query

if (! rs.next()) // no rows found?

System.out.println("no authorization for this username...");
else
{
String ts = rs.getString("TS"); // retrieve the timestamp (key)
String hostval = rs.getString("HOST"); // retrieve allowed hostname

270 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

// write the current connection info into a table

// with this table it is possible to use the SQL like function String
insertStr = "insert into APPLACCESS values " +
"('"+DB2USR+"','"+ipname+"','"+ipaddr+"','"+ts+"')";
System.out.println("Insert = " + insertStr);
PreparedStatement ps0 = db2Conn.prepareStatement(insertStr);
ps0.execute(); // run the insert
// check if the current connection info has an equivalent host entry
// (either IP-name or IP-address)
String query2 = "select 1 from APPLACCESS where "+
"TS='"+ts+"' and "+
"HOSTNAME like '"+hostval+"' or IPADDR like '"+hostval +"'";
while (rs.next()) // there can be more than one permitted hosts
{
hostval = rs.getString("HOST"); // retrieve the allowed hostname
query2 = query2 + " or HOSTNAME like '" + hostval +
"' or IPADDR like '" + hostval + "'";
}
System.out.println("Query = " + query2);
PreparedStatement ps2 = db2Conn.prepareStatement(query2);
ResultSet rs2 = ps2.executeQuery(); // run the query
if (! rs2.next()) // no accordance found?
System.out.println("no authorization...");
else
{
System.out.println("You are authorized to go on!");
// here should be the call to the access controlled function if (mode.equals("SEL"))
{
// call the SELECT function
}
if (mode.equals("INS")
{
// call the INSERT function
}
}
}
}
}

This example application works in the following ways:

򐂰 The first step is to get all hosts out of the ACCESSLIST table from which the
specified user has access to the requested function.
򐂰 The second step is to insert the information about the access request into the
table APPLACCESS. The main reason for this step is that if this information is
stored in a table, the SQL LIKE function can be used in the next step. The
LIKE function handles wildcards (“%” and “_”) in the host information
correctly.
򐂰 The third step is to verify if the host name or IP address has access rights by
comparing the entry with the host name retrieved in the first step.

Chapter 8. Application conversion 271

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 If access is allowed, the function should be executed. You can also

implement a method that has a return code telling if access is allowed or not.

8.3 Additional application considerations

Once converted to DB2 9.7, an application’s run-time performance can be
impacted by a number of factors. This section describes what locking and
transaction isolation does to your application when running in a multi-user
environment.

8.3.1 What is the purpose of locking?

When many users access the same data source using your application, or any
other interface that allows data manipulation, some unwanted effects may occur.
These effects are called:
򐂰 Lost update:
Two concurrent users retrieve and update the same data. The last successful
change is kept while the first change is overridden.
򐂰 Uncommitted (or dirty) read:
User A can read or view data changed by User B, but those changes have not
been committed yet.
򐂰 Non-repeatable read:
Within the same transaction User A runs the identical SELECT statement
multiple times with different results because User B modified records in User
A’s result set.
򐂰 Phantom read:
Within the same transaction User A runs a SELECT statement multiple times
and gets addition records because user B added records in User A’s result
set.

One of the more advanced features of a data management system is to define

modification rules to control the use of data and guarantee the integrity of the
data to prevent the above undesirable effects.

8.3.2 Concurrency control and transaction isolation

From a bird’s eye view, two methods for concurrency control can be
differentiated:

272 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

򐂰 The optimistic concurrency approach:

A strategy to increase concurrency in which rows are not locked.
Transactions are divided into read, validate, and write phases. Instead, during
the validation phase before they are updated or deleted, a cursor checks to
see if they have been changed since they were last read. If so, the update or
delete fails.
򐂰 The pessimistic concurrency or locking approach:
A strategy in which each rows are locked so that other transactions cannot
change them. Transaction requests locks to update resources. Other
transactions have to wait or time-out. The resource is released on transaction
completion or commit and rollback.

Both methods have their pros and cons, but by far the most popular method is
the latter approach. Both MySQL and DB2 follow this approach to various
degrees of sophistication and implementation differences.

8.3.3 DB2 isolation levels

The isolation level that is associated with an application process determines the
degree to which the data that is being accessed by that process is locked or
isolated from other concurrently executing processes. The isolation level is in
effect for the duration of a unit of work.

The isolation level of an application process therefore specifies:

򐂰 The degree to which rows that are read or updated by the application are
available to other concurrently executing application processes.
򐂰 The degree to which the update activity of other concurrently executing
application processes can affect the application.

The isolation level for static SQL statements is specified as an attribute of a

package and applies to the application processes that use that package. The
isolation level is specified during the program preparation process by setting the
ISOLATION bind or precompile option. For dynamic SQL statements, the default
isolation level is the isolation level that was specified for the package preparing
the statement. Use the SET CURRENT ISOLATION statement to specify a
different isolation level for dynamic SQL statements that are issued within a
session. For both static SQL statements and dynamic SQL statements, the
isolation-clause in a select-statement overrides both the special register (if set)
and the bind option value.

Because the isolation level determines how data is isolated from other processes
while the data is being accessed, you should select an isolation level that

Chapter 8. Application conversion 273

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

balances the requirements of concurrency and data integrity. Table 8-5 on

page 274 gives an overview over DB2 isolation levels.

Table 8-5 DB2 isolation level

DB2 isolation level Description

Uncommitted Read 򐂰 Access to uncommitted data from other transaction

򐂰 No record locks unless updates occur

Cursor Stability 򐂰 Addresses dirty read issue

򐂰 Sees only committed data from other transaction
򐂰 Lock is only held on cursor position unless update occurs
򐂰 Update lock is held until transaction completed = commit
򐂰 default isolation level

Read Stability 򐂰 Addresses non-repeatable read issue

򐂰 Sees only committed data from other transaction
򐂰 Locks are held on every row fetched (Inserts permitted)
򐂰 Locks are held for duration of transaction
(commit/rollback)

Repeatable Read 򐂰 Addresses phantom read issue

򐂰 All record locks held for duration of transaction
򐂰 A repeated query within the same transaction will get the
same result set (Inserts are prevented)

Note: For Cursor Stability isolation level, the currently committed semantics
introduced in Version 9.7, only committed data is returned. This was the case
in previous releases, but now readers do not wait for updaters to release row
locks. Instead, readers return data that is based on the currently committed
version; that is, data prior to the start of the write operation.

As one can see, the isolation levels listed in Table 8-5 are ordered descendent
according to the number and duration of locks held during the transaction, and
therefore the degree of concurrency or locking is required to ensure the desired
level of data integrity. However, as we can see too much locking drastically
reduces concurrency. Poor application design and coding may cause locking
problems such as:
򐂰 Deadlocks
򐂰 Lock waits
򐂰 Lock escalation
򐂰 Lock time-outs

By default DB2 operates with the isolation level cursor stability. Transaction
isolation can be specified at many different levels as discussed in 8.3.5,

274 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

“Specifying the isolation level in DB2” on page 277. For good performance, verify
the lowest isolation level required for your converted application.

For additional information on the DB2 concurrency implementation, refer to the

IBM DB2 Database for Linux, UNIX, and Windows Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

8.3.4 Locking
Some MySQL applications when ported to DB2 appear to behave identically, and
the topic of concurrency can be ignored. However, if your applications involve
frequent access to the same tables you may experience a different behavior. By
default, MySQL runs in a mode that is called autocommit. This means that
MySQL considers each and every SQL statement as an atomic unit of work or
transaction.

In contrast, DB2 by default considers a group of SQL statements with the

corresponding unit of work boundaries set by a commit, a rollback statement as
single or atomic transaction, respectively. There are certain interfaces such as
the DB2 command line processor (CLP) or the JDBC interface that run in
autocommit mode. For other application interfaces, autocommit is turned off by
default.

Another matter causing heated discussions among experts is the level of locking
required for implementation on the database level. Should the locking approach
be implemented with the lowest level of overhead, and therefore maintain locks
on a table level? Or, is it better to lock on a lower level for example on page
level? Should the granularity be even finer and locking occur on row level?

As usual, the correct answer to these questions is: it depends!

MySQL development decided to go the multi storage engine way and decided to
implement lock levels based on the type of table. Table types can be mixed
within a database and even a statement, and types can be altered. The default
storage engine for MySQL supports only table level locking. MyISAM table,
Merge and MEMORY tables use table level locking. The InnoDB storage engine
was released as a transactional table handler of MySQL with a lock manager for
row level locking mechanisms. Hence, the MySQL table type InnoDB defines
tables, most alike DB2 tables. Table 8-6 gives a superficial comparison of the
different flavors of MySQL tables with DB2 tables:

Chapter 8. Application conversion 275

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

Table 8-6 MySQL and DB2 table comparison

Characteristics DB2 tables MyISAM tables InnoDB tables

Lock level Row level, None or table level Row level and table
Table level only on level
explicit request

Commitment Yes No Yes

control

Isolation level Uncommitted No Read

Reads, Cursor Uncommitted,
Stability, Read Read Committed,
Stability, Repeatable Reads,
Repeatable Reads Serializable

Rollback on DDL Yes No Yes

Consider using table level locking for:

򐂰 Applications that use mostly reads such as Data Warehouse and Business
Intelligence applications
Applications reading and updating through key positioning such as
UPDATE... WHERE Custno =?
򐂰 Applications using INSERTs with subselects and only a small number of
UPDATE and DELETEs

Consider to use row level locking for:

򐂰 Applications that require a high level of concurrency and OLTP capabilities
򐂰 Many SELECTs with only small result sets
򐂰 Applications with high UPDATE/INSERT/DELETE frequency

However, let us attempt to summarize the concurrency issues that may arise
when converting a MySQL application to DB2 based on the two MySQL table
types, which we consider significant:
򐂰 MyISAM tables provide a high level of concurrency since SQL processing
occurs in autocommit mode and no row level locks are maintained. When
converting to DB2 ensure your application operates in autocommit mode.
Verify the lowest isolation level required for your application and MyISAM
tables.
򐂰 InnoDB tables provide concurrency control very similar to DB2. Note that
default transaction isolation for InnoDB is repeatable read..

276 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

8.3.5 Specifying the isolation level in DB2

Because the isolation level determines how data is locked and isolated from
other processes, while it is being accessed, you should select an isolation level
that balances the requirements of concurrency and data integrity.

The isolation level that you specify is in effect for the duration of the unit of work.
The isolation level can be specified in several different ways. The following
heuristics are used to determine which isolation level will be used when
compiling an SQL or XQuery statement:
򐂰 Static SQL:
– If an isolation clause is specified in the statement, then the value of that
clause is used.
– If no isolation clause is specified in the statement, then the isolation level
used is the one specified for the package at the time when the package
was bound to the database.
򐂰 Dynamic SQL:
– If an isolation clause is specified in the statement, then the value of that
clause is used.
– If no isolation clause is specified in the statement, and a SET CURRENT
ISOLATION statement has been issued within the current session, then
the value of the CURRENT ISOLATION special register is used.
– If no isolation clause is specified in the statement, and no SET CURRENT
ISOLATION statement has been issued within the current session, then
the isolation level used is the one specified for the package at the time
when the package was bound to the database.
򐂰 For static or dynamic XQuery statements, the isolation level of the
environment determines the isolation level that is used when the XQuery
expression is evaluated.

Note: Many commercially written applications provide a method for choosing

the isolation level. Refer to the application documentation for information.

SQL procedure and isolation level

This section discusses when to specify the isolation level for a SQL procedure.

At precompile or bind time

For an application written in a supported compiled language, use the
ISOLATION option of the Command Line Processor PREP or BIND commands.
You can also use the sqlaprep or sqlabndx API to specify the isolation level.

Chapter 8. Application conversion 277

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 If you create a bind file at precompile time, the isolation level is stored in the
bind file. If you do not specify an isolation level at bind time, the default is the
isolation level used during precompilation.
򐂰 If you do not specify an isolation level, the default of cursor stability is used.

Tip: To determine the isolation level of a package, execute the following

query:
SELECT ISOLATION FROM SYSCAT.PACKAGES
WHERE PKGNAME = 'XXXXXXXX'
AND PKGSCHEMA = 'YYYYYYYY'

where XXXXXXXX is the name of the package and YYYYYYYY is the schema
name of the package. Both of these names must be in all capital letters.

On database servers that support REXX

When a database is created, multiple bind files that support the different isolation
levels for SQL in REXX are bound to the database. Other command-line
processor packages are also bound to the database when a database is created.

REXX and the command line processor connect to a database using a default
isolation level of cursor stability. Changing to a different isolation level does not
change the connection state.

To determine the isolation level that is being used by a REXX application, check
the value of the SQLISL predefined REXX variable. The value is updated each
time that the CHANGE ISOLATION LEVEL command executes.

At the statement level

Use the WITH clause. The WITH clause cannot be used on subqueries. The
WITH UR option applies to read-only operations only. In other cases, the
statement is automatically changed from UR to CS.

This isolation level overrides the isolation level that is specified for the package in
which the statement appears. You can specify an isolation level for the following
SQL statements:
򐂰 DECLARE CURSOR
򐂰 Searched DELETE
򐂰 INSERT
򐂰 SELECT
򐂰 SELECT INTO
򐂰 Searched UPDATE

278 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch08.fm

Note: Isolation levels for XQuery statements cannot be specified at the

statement level.

From CLI or ODBC at runtime

Use the CHANGE ISOLATION LEVEL command. With DB2 Call-level Interface
(CLI), you can change the isolation level as part of the CLI configuration. At run
time, use the SQLSetConnectAttr function with the
SQL_ATTR_TXN_ISOLATION attribute to set the transaction isolation level for
the current connection referenced by the ConnectionHandle argument. You can
also use the TXNISOLATION keyword in the db2cli.ini file

When working with JDBC or SQLj at runtime

To create a package (and specify its isolation level) in SQLJ, use the SQLJ
profile customizer (db2sqljcustomize command).

Note: JDBC and SQLj are implemented with CLI on DB2, which means the
db2cli.ini settings might affect what is written and run using JDBC and SQLj.

For dynamic SQL within the current session

Use the SET CURRENT ISOLATION statement to set the isolation level for
dynamic SQL issued within a session. Issuing this statement sets the CURRENT
ISOLATION special register to a value that specifies the isolation level for any
dynamic SQL statements that are issued within the current session. Once set,
the CURRENT ISOLATION special register provides the isolation level for any
subsequent dynamic SQL statement that is compiled within the session,
regardless of which package issued the statement. This isolation level is in effect
until the session ends or until the SET CURRENT ISOLATION…RESET
statement is issued.

Chapter 8. Application conversion 279

7093ch08.fm Draft Document for Review October 16, 2009 10:22 am

280 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

Chapter 9. Database administration

In this chapter, we focus on the database administration features that are offered
by DB2. We provide a general introduction and detailed description to DB2
administration programs, utilities, and tools. We also cover a few of the salient
features available in DB2 but missing in MySQL.

We present key attributes of database administration for DB2 such as:

򐂰 Database configuration
򐂰 Data recovery
򐂰 Database replication
򐂰 Data movement utilities
򐂰 High availability
򐂰 Autonomics
򐂰 Workload Manager

We also explore the following graphical tools:

򐂰 DB2 Control Center
򐂰 IBM Data Studio
򐂰 Data Studio Administration Console

© Copyright IBM Corp. 2009. All rights reserved. 281

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

9.1 Database configuration

Database configuration is a very important task for database administrators. It
involves setting up the optimal parameters for the database, system, database
manager and all other related components to achieve the best performance for
your application. In this section we discuss how DB2 database parameters are
tuned.

9.1.1 DB2 configuration

DB2 has two levels of configuration:
򐂰 Database manager (instance)
򐂰 Database

DB2 environment variables and profile registry

DB2 environment and profile registry variables control the database
environment. Information stored in these profile registries is used by the DB2
server instance and any DB2 applications started after the changes have been
made.

Using the DB2 profile registry allows for centralized control of the environment
variables. Through use of different profiles, different levels of support are
provided. Remote administration of the environment variables is also available
when using the DB2 Administration Server.

There are four profile registries:

򐂰 The DB2 instance level profile registry
The majority of the DB2 environment variables are placed within this registry.
The environment variable settings for a particular instance are kept in this
registry. Values defined in this level override their settings on the global level.
򐂰 The DB2 global level profile registry
If an environment variable is not set for a particular instance, this registry is
used. This registry is visible to all instances pertaining to a particular copy of
DB2 data server, one global-level profile exists in the installation path.
򐂰 The DB2 instance node level profile registry
This registry level contains variable settings that are specific to a database
partition in a partitioned database environment. Values defined in this level
override their settings at the instance and global levels.
򐂰 The DB2 instance profile registry

282 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

This registry contains a list of all instance names associated with the current
copy. Each installation has its own list. You can see the complete list of all
instances available on the system by running db2ilist.

The variables can be set using the db2set command. The command immediately
stores the updated variables in the profile registry. Example 9-1 shows different
modes in which the db2set command can be used.

Example 9-1 Changing registry and environment variables using db2set

//this shows the current registry variables//
db2inst1@db2server:~> db2set -all
[i] DB2_COMPATIBILITY_VECTOR=8
[i] DB2PROCESSORS=0,1
[i] DB2COMM=tcpip
[g] DB2SYSTEM=db2server
[g] DB2INSTDEF=db2inst1
[g] DB2ADMINSERVER=dasusr1

//this sets registry for all databases in particular instance//

db2inst1@db2server:~>db2set DB2AUTOSTART=YES -i db2inst1

//this sets registry variable for all instances//

db2inst1@db2server:~>db2set DB2AUTOSTART=YES -g

//this sets registry variable for particular node//

db2inst1@db2server:~>db2set DB2AUTOSTART=YES -i db2inst1 65000

DB2 configures the operating environment by checking for registry values and
environment variables, and resolves them in the following order:
1. Environment variables are set using the set command. (Or the export
command on UNIX platforms.)
2. Registry values set with the instance node-level profile (using the db2set -i
<instance name> <nodenum> command).
3. Registry values set with the instance-level profile (using the db2set -i
command).
4. Registry values set with the global-level profile (using the db2set -g
command).

DB2 Configuration parameters

In addition to DB2 profile registry, DB2 also has instance and database
configuration files, which provide users with flexibility to configure the database
and the database engine to fit business and application needs. These files
contain parameters that define values such as resources allocated to DB2,
diagnostic level, log files location, and others.

Chapter 9. Database administration 283

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

When a DB2 database instance or a database is created, a corresponding

configuration file is created with default parameter values. It is strongly
suggested to modify these parameters to improve performance of the instance or
database.

Figure 9-1 illustrates the two DB2 configuration files and additional operating
system configurations.

Figure 9-1 Configuration parameter files

򐂰 The database manager configuration file

This file is created automatically when creating a DB2 instance and affects
the configuration on the instance level. It is stored in db2systm file under the
sqllib subdirectory of instance. Use the following command to see the values
in this file:
db2 GET DATABASE MANAGER CONFIGURATION
and parameter values can be set using:
db2 UPDATE DBM CFG USING <parameter> <value>
򐂰 The database configuration file
For each created database, a database configuration file is created as well. It
is stored in a file named SQLDBCONF and resides in the directory where the
database resides. It defines the resources and variables for the particular
database. Use the following command to obtain the values in this file:
db2 GET DATABASE CONFIGURATION FOR <database name>
and values can be set using:

284 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

db2 UPDATE DATABASE CONFIGURATION FOR <database name> USING <parameter>

<value>

Configuration tools
IBM has tools to assist with configuring your database server. Two of which are
the Configuration Assistant and the IBM Data Studio.

You can use the DB2 Configuration Assistant to configure and maintain the
database objects that you or your application will be using. The Configuration
Assistant is a graphical tool tightly integrated with the DB2 Control Center. It
allows you to update both the DB2 Profile Registry and the DB2 database
manager configuration parameters on the local machine as well as remotely. It
can be launched from the Control Center or by calling the db2ca utility. The
Configuration Assistant also has an advanced view, which uses a notebook to
organize connection information by object: Systems, Instance Nodes,
Databases, Database Connection Services (DCS), and Data Sources. Figure 9-2
shows how to change the database manager configuration using the
Configuration Assistant.

Figure 9-2 DB2 Configuration Assistan

Chapter 9. Database administration 285

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

Note: The Configuration Assistant has been deprecated in Version 9.7 and
might be removed in a future release. We recommend the IBM Integration
Management solutions for managing DB2 for Linux, UNIX, and Windows data
and data-centric applications.

The IBM Data Studio is part of the IBM Integrated Management solutions for
managing your DB2 database. Data Studio simplifies the process of managing
your database objects by supporting instance and database management and by
providing the ability to run database commands and utilities. It provides a simple
user interface to invoke the database administration commands that you use to
maintain and manage your database environment. Figure 9-3 shows how to
change the database manager and database configuration using Data Studio.

Figure 9-3 Data Studio- Configuration window

IBM offers a number of automatic tools and DB2 features to make database
administration effortless. The Configuration Advisor can be used to assist with
parameter configuration and configuring your database for optimal performance.
The Configuration Advisor looks at your current database, asks for user input on

286 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

the database workload and suggests the best configuration parameters for buffer
pool size, database configuration, and database manager configuration.
Figure 9-4 shows the suggested output for our sample inventory database.

Figure 9-4 Configuration Advisor

9.2 Database recovery

Database recovery is the action the database system or user takes to recuperate
the database in case of system, hardware, software, or application failure. This
includes precautionary measures and curing measures. Different and multiple
recovery methods are followed by all the database systems for warding off the
data lose condition.

9.2.1 DB2 database recovery

DB2 recovery method is much more sophisticated then MySQL and allows you to
build an optimized backup and recovery strategy for your database system. DB2
supports database recovery using database backup in conduction with three
recovery logs shown in Figure 9-5 on page 288. The recovery logs files and the
recovery history files are created automatically when a database is created.

Chapter 9. Database administration 287

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

These log files are important if you need to recover data that is lost or damaged.
The three files work in the following ways:
򐂰 Recovery log files
The recovery log files are used to recover from application or system errors.
In combination with the database backups, they are used to recover the
consistency of the database right up to the point in time when the error
occurred. There are recovery log files for each database on the server.
򐂰 Recovery history files
The recover history files contain a summary of the backup information that
can be used to determine recovery options, if all or part of the database must
be recover to a given point in time. It is used to track recovery-related events
such as backup and restore operations, among others. This file is located in
the database directory.
򐂰 Table space change history file
The table space change history files are also located in the database
directory. These files contain information that can be used to determine which
log files are required for the recover of a particular table space.

Figure 9-5 Database recovery files

There are two types of databases for backup and recovery, non-recoverable and
recoverable databases. As a database administrator you need to decide which
category your database falls under.
򐂰 Non-recoverable database
Data that is easily recreated can be stored in a non-recoverable database.
This includes data that is used for read-only applications and tables that are
not often updated. These types of databases have a small amount of logging
which does not justify the added complexity of managing log files and rolling
forward after a restore operation.

288 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

To setup a non-recoverable database set both database configuration

parameters logarchmeth1 and logarchmeth2 to OFF. This means that only
the crash recovery logs are kept. These logs are known as active logs, and
they contain current transaction data. Because non-recoverable databases
only support active logs they do not support roll forward recovery.
򐂰 Recoverable database
Data that cannot be easily recreated should be stored in a recoverable
database. This includes data whose source is destroyed after the data is
loaded, data that is manually entered into tables, and data that is modified by
application programs or users after it is loaded into the database.
To setup a recoverable database set both logarchmeth1 and logarchmeth2
database configuration parameters to anything other than OFF. On top of the
active logs kept for crash recovery recoverable databases also have archived
logs. These logs contain committed data and therefore support crash, version
or roll forward recovery. Recoverable databases also support backup,
restore, and roll forward individual tables spaces.

Database backup
To back up a DB2 database, database partition, or a selected table space, you
can use the DB2 backup command. This command can be used to create a
backup to disk, tape, or named pipes in UNIX. DB2 supports both offline and
online backup.
db2 backup database invent to /home/db2inst1/backup

You can back up an entire database, database partition, or only selected table
spaces.

In addition to backing up the entire database every time, DB2 also supports
incremental backups where you can back up large databases on a regular basis
incrementally. This requires a trackmod database configuration parameter to be
set to yes. Incremental backup can be a cumulative backup, which stores data
changes since the last successful full backup, or delta backup which is the last
successful backup irrespective of whether that backup was full, delta, or
cumulative. Figure 9-6 on page 290 and Example 9-2 on page 290 show the
cumulative and delta backup technique.

Chapter 9. Database administration 289

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

Figure 9-6 Incremental backup

Example 9-2 Incremental backup

db2inst1@db2server:~> db2 BACKUP DATABASE invent TO /home/db2inst1/backup
db2inst1@db2server:~> db2 BACKUP DATABASE invent INCREMENTAL TO
/home/db2inst1/incBackup
db2inst1@db2server:~> db2 BACKUP DATABASE invent INCREMENTAL DELTA TO
/home/db2inst1/deltaBackup

Since it can be time-consuming to determine when it is required and whether

maintenance activities - such as backup operations - should run, IBM has tools to
assist you with this confirmation. You can use the Configure Automatic
Maintenance wizard within Data Studio, shown in Figure 9-7 on page 291, or the
DB2 Control Center to configure the database maintenance activities. With
automatic maintenance, you specify your maintenance objectives, including
when automatic maintenance can run. DB2 then uses these objectives to
determine if the maintenance activities need to be done and then runs only the
required maintenance activities during the next available maintenance window (a
user-defined time period for the running of automatic maintenance activities).

290 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

Figure 9-7 Data Studio- Configuration Automatic Maintenance

Database recovery
The recover utility performs the necessary restore and roll forward operations to
recover a database to a specified time, based on information found in the
recovery history file. When you use this utility, you specify that the database be
recovered to a point-in-time or to the end of the log files. The utility will then
select the best suitable backup image and perform the recovery operations.
Example 9-3 shows how to use the RECOVER DATABASE command.

Example 9-3 Recover Database

b2inst1@db2server:~> db2 RECOVER DB invent

Rollforward Status

Input database alias = invent

Number of nodes have returned status = 1

Node number = 0
Rollforward status = not pending
Next log file to be read =
Log files processed = S0000000.LOG - S0000001.LOG
Last committed transaction = 2009-09-11-19.25.01.000000 Local

DB20000I The RECOVER DATABASE command completed successfully.

Chapter 9. Database administration 291

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

There are three types of recovery:

򐂰 Crash Recovery
Crash recovery protects a database from being left in an inconsistent or
unusable state, which occurs when transactions are interrupted
unexpectedly. This is done by rolling back incomplete transactions and
completing committed transactions that were still in memory when the crash
occurred.
This is done automatically by the database manager if you set the automatic
restart (autorestart) database configuration parameter to ON by:
db2inst1@db2server:~> db2 UPDATE DATABASE CONFIGURATION FOR invent USING
AUTORESTART ON
or by restarting the database when a database failure occurs; this is done by
calling:
db2inst1@db2server:~> db2 RESTART DATABASE invent
DB2 maintains log files, the recovery history file, and table space change
history file to recover data that is lost or damaged.
򐂰 Version Recovery
Version recovery is the restoration of a previous version of the database,
using an image that was created during a backup operation. You can use this
recovery option for non-recoverable database (databases without archived
logs). You can also use this method with recoverable databases by using the
WITHOUT ROLLING FORWARD option in the RESTORE DATABASE
command. A database restore operation will restore the entire database using
a backup image created earlier. A database backup allows you to restore a
database to a state identical to the one at the time that the backup was made.
However, every unit of work from the time of the backup to the time of the
failure is lost, as shown in Figure 9-8 on page 293.
It is very necessary to take a backup image on a regular basis because this
recovery method loses the changes made in the database after the backup
operation. This is done by restoring the database using incremental, delta, or
full backup.

292 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

Figure 9-8 Version Recovery

򐂰 Roll Forward Recovery

Roll forward recovery can be used to reapply changes that were made by
transactions that were committed after a backup was made.
There are two types of roll forward recovery:
– Database roll forward recovery
With this type of roll forward recovery, you can tell DB2 to roll forward a
database to the state immediately before the failure or you can specify the
local time to which you want to roll forward your database. Figure 9-9
shows the roll forward recovery technique of a DB2 database.

Chapter 9. Database administration 293

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

Figure 9-9 Version Recovery - Database roll forward recovery

– Table space roll forward recovery

With this type of roll forward recovery, you can tell DB2 to roll forward a
table space to either a state immediately before the failure or to a
particular point in time. Figure 9-10 shows the roll forward recovery
technique of a DB2 table space.

Figure 9-10 Version Recovery - Table space roll forward recovery

Figure 9-11 shows how you can use Data Studio to recover a DB2 database.

294 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

Figure 9-11 Data Studio- Recover Option

Database restore
DB2 database restore is as easy as taking database backup. This can be done
by using the RESTORE utility. The restore database command rebuilds the
database data or table space to the state that it was in when the backup copy
was made. This utility can overwrite a database with a different image or restore
the backup copy to a new database. The restore utility can also be used to
restore backup images in DB2 Version 9.7 that were backed up on DB2
Universal Database Version 8, DB2 Version 9.1, or DB2 Version 9.5.

This RESTORE utility supports full and incremental database restore. Incremental
database restore can be automatic or manual. Example 9-4 shows automatic
incremental restore, and Example 9-5 shows manual incremental restore.

Example 9-4 Automatic incremental restore

db2inst1@db2server:~ > db2 RESTORE DATABASE invent INCREMENTAL AUTOMATIC FROM
/home/db2inst1/backup TAKEN AT 20090911214318

Example 9-5 Manual incremental restore

db2inst1@db2server:~ > db2 RESTORE DATABASE invent INCREMENTAL FROM
/home/db2inst1/backup TAKEN AT 20090909

Chapter 9. Database administration 295

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

db2inst1@db2server:~ > db2 RESTORE DATABASE invent INCREMENTAL FROM

/home/db2inst1/backup TAKEN AT 20090910
db2inst1@db2server:~ > db2 RESTORE DATABASE invent INCREMENTAL FROM
/home/db2inst1/backup TAKEN AT 20090911

If at the time during backup operation the database was enabled for roll forward
recovery, the database can be brought to its previous state by invoking the
following ROLLFORWARD command after successful completion of a restore
operation:
db2inst1@db2server:~ > db2 ROLLFORWARD DATABASE invent COMPLETE

Tip: DB2 backup and restore can also be performed using the Data Studio

The RESTORE and ROLLFORWARD utilities can also be executed from Data Studio.
Figure 9-12 on page 296 shows the restore database window.

Figure 9-12 Data Studio- Restore option

296 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

9.3 Database replication

Database replication is the process of maintaining the same database under
multiple servers or systems. It involves synchronizing changes from one
database (a source) to another database (a target). This feature is very useful for
load sharing, fast disaster recovery, and high availability.

IBM provides two different solutions that you can use to replicate data from and
to relational databases: SQL replication and Q replication. IBM also provides a
solution called event publishing for converting committed source changes into
messages in an XML or delimited format and for publishing those messages
across WebSphere MQ queues to applications.

In SQL replication, committed source changes are staged in relational tables

before being replicated to target systems. Q replication is a replication solution
that can replicate large volumes of data at low levels of latency. In Q replication,
committed source changes are written in messages that are transported through
WebSphere MQ message queues to target systems.

Replication is not only supported between two DB2 systems running on different
platforms, but also on the following non-DB2 databases SQL replication supports
replicating between DB2 on Linux, UNIX, Windows, z/OS, and iSeries; Informix;
Microsoft SQL Server; Oracle; Sybase; and Teradata (target only). Q replication
supports DB2 for Linux, UNIX, and Windows, DB2 for z/OS, Informix (target
only), Microsoft SQL Server (target only), Oracle (target only), and Sybase
(target only).

IBM provides three tools that can assist with setting up replication.
򐂰 Replication Center
The Replication Center is a graphical user interface that you can use to
define, operate, and monitor your replication and publishing environments. It
comes with the DB2 Administration Client and runs on Linux and Windows
systems. The Replication Center provides a single interface for administering
your replication environments on different platforms across multiple systems.
Among its features:
– Launch pads that show you step by step how to configure basic replication
and publishing environments.
– Wizards that help you set up simple to highly customized replication and
publishing configurations.
– Profiles that you can customize that let you create replication objects with
schemas, names, and other attributes that conform to your own
conventions and storage requirements.

Chapter 9. Database administration 297

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

The Replication Center can be invoked through Control Center or using db2rc
command.
򐂰 ASNCLP command-line program
The ASNCLP program generates SQL scripts for defining and changing
replication and publishing environments. The program runs on Linux, UNIX,
Windows, and UNIX System Services (USS) for z/OS. The ASNCLP program
does not run natively on z/OS or System i.
You can use the ASNCLP to administer SQL replication, Q replication,
Classic replication, event publishing, and the Replication Alert Monitor. You
can build ASNCLP input files and run them to generate SQL scripts, or you
can run ASNCLP commands interactively from an operating system prompt.
You can also run the ASNCLP program in execute-immediately mode, which
is useful for operational commands such as START QSUB, STOP QSUB, or
LIST.
򐂰 Replication Dashboard
The Replication Dashboard is a Web-based, graphical user interface that
helps you monitor and manage the health of replication and event publishing.
The dashboard provides an overall summary of replication and publishing
configurations in a convenient tabular format with high-level status indicators.
You can drill down for more detailed information on queues and queue depth,
subscriptions, latency, and exceptions, and generate detailed reports to help
track performance or diagnose problems. You can also view up to 24 moving
graphs for near-real-time performance information.
You can change program parameters, start, stop, and reinitialize
subscriptions, and start and stop queues. The dashboard also provides a
convenient way to view alerts from the Replication Alert Monitor program.

9.4 Data movement

In this section we discuss data movement support by DB2. The data movement
utilities are used to transfer data between different tables, databases, or
database systems.

9.4.1 DB2 data movement

On the contrary, DB2 provides faster and efficient tools and utilities for data
movement across the different systems or reorganizing data on the same
system.

298 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

EXPORT utility
DB2 EXPORT is a powerful tool to export your DB2 data quickly from DB2 to the
external file system. DB2 EXPORT uses SQL select or an XQuery statement to
export tables, views, large objects, or typed tables to one of the three external file
formats:
– .DEL: delimited ASCII format file
– .WSF: work sheet format like lotus 1-2-3®
– .IXF: integrated exchange format

The EXPORT utility can be invoked through:

򐂰 The Command Line Processor (CLP)
The EXPORT utility can be used from CLP by supplying an SQL SELECT or
XQuery statement, or by providing hierarchical information for typed tables as
shown below:
db2inst1@db2server:~ > db2 EXPORT TO invent.ixf of ixf SELECT * FROM
admin.owners
򐂰 Data Studio
Data from a DB2 database can be extracted using the graphical user
interface - Data Studio. This tool allows you to set export options for each
table visually. Figure 9-13 shows the usage of the export through Control
Center.

Chapter 9. Database administration 299

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

Figure 9-13 Data Studio- Export Option

򐂰 The Control Center

Another graphical user interface that can be used is the Control Center.
򐂰 ADMIN_CMD stored procedure
The ADMIN_CMD can be used by applications to run administrative calls,
such as the export command.
򐂰 An application programming interface (API) - db2Export
DB2 provides an API for exporting data, which can be used to export data
programmatically. This can be used by importing db2ApiDf.h and the method
db2Export.

300 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

The IMPORT utility

The files created with the same syntax as the EXPORT utility can be used to
populate data into a new DB2 database on the same system, or they can be
transferred to another platform and imported or loaded to the DB2 database that
resides on that platform. The IMPORT utility supports the following file formats:
– ASC: non-delimited ASCII format file
– .DEL: delimited ASCI format file
– .WSF: work sheet format like lotus 1-2-3
– .IXF: integrated exchange format

As EXPORT, IMPORT utility can also be used from:

򐂰 The Command line processor (CLP)
The IMPORT utility can be used from CLP by supplying an SQL INSERT,
INSERT UPDATE, or REPLACE option. The example below shows the
simple IMPORT statement:
db2inst1@db2server:~ > db2 IMPORT FROM invent.ixf OF ixf MESSAGES
msg.txt INSERT INTO admin.owners
򐂰 Data Studio
Data Studio can be used to import data graphically. Figure 9-14 on page 302
shows the import wizard.

Chapter 9. Database administration 301

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

Figure 9-14 Data Studio- Import Option

򐂰 The Control Center

Control Center can also be used for importing data visually.
򐂰 ADMIN_CMD stored procedure
The ADMIN_CMD can be used by applications to run administrative calls and
run the IMPORT command.
򐂰 An application programming interface (API) - db2Import
DB2 provides an API for importing data from files such as files exported using
the export tool. This provides an option to do importing programmatically by
using db2ApiDf.h and a method db2import.

The DB2MOVE utility

Data can be copied using the DB2 EXPORT and IMPORT utilities. But a more
efficient way to copy an entire DB2 database schema is by using the DB2

302 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

db2move utility. This utility queries the system catalog tables for a specified
database, and exports the table structure and contents of each table found to a
PC/IXF formatted file. These files can be used to populate another DB2
database. It can be run in three modes:
򐂰 EXPORT mode
In this mode, the db2move utility invokes DB2 EXPORT utility to extract data from
one or more tables and write to PC/IXF formatted files. It also creates a file
named db2move.lst that contains the names of all tables exported and the
names of the files that the table data was written to. EXPORT mode can be
used as below:
db2inst1@db2server:~ > db2move invent EXPORT
򐂰 IMPORT mode
In this mode, the db2move utility invokes the DB2 IMPORT utility to recreate
tables and indexes from data stored in PC/IXF formatted files. The file
db2move.lst generated in EXPORT mode can be used to get information
about tables in the exported files. The above exported files can be imported
using:
db2inst1@db2server:~ > db2move invent IMPORT
򐂰 LOAD mode
In this mode, the db2move utility invokes the DB2 LOAD utility to populate
tables that already exist with data stored in PC/IXF formatted files. The file
db2move.lst generated in EXPORT mode can be used to get information
about tables. The above exported files can be loaded using:
db2inst1@db2server:~ > db2move invent LOAD -l /home/db2inst1/export

The LOAD utility

The LOAD utility can be used for moving large amounts of data into a newly
created table, or into a table that already contains data. The utility can handle
most data types, including XML, large objects (LOBs), and user-defined types
(UDTs). The LOAD utility is faster than the IMPORT utility, because it writes
formatted pages directly into the database, while the IMPORT utility performs SQL
INSERTs. The LOAD utility does not fire triggers, and does not perform referential
or table constraints checking (other than validating the uniqueness of the
indexes). The LOAD utility supports the following import sources.
򐂰 ASC: non-delimited ASCII format file
򐂰 .DEL: delimited ASCI format file
򐂰 .IXF: integrated exchange format

There are four different modes you can execute the LOAD utility in:
򐂰 INSERT

Chapter 9. Database administration 303

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

In this mode, load appends input data to the table without making any
changes to the existing data.
򐂰 REPLACE
In this mode, load deletes existing data from the table and populates it with
the input data.
򐂰 RESTART
In this mode, an interrupted load is resumed. In most cases, the load is
resumed from the phase it failed in. If that phase was the load phase, the load
is resumed from the last successful consistency point.
򐂰 TERMINATE
In this mode, a failed load operation is rolled back.

The LOAD utility can also be used from:

򐂰 The Command line processor (CLP)
The IMPORT utility can be used from CLP by supplying an SQL INSERT,
INSERT UPDATE, or REPLACE option. The example below shows the
simple IMPORT statement:
db2inst1@db2server:~ > db2 LOAD FROM ownersdata.del OF del REPLACE INTO
admin.owner
򐂰 Data Studio
Data Studio can be used to load data graphically. Figure 9-15 on page 304
shows the import wizard.

Figure 9-15 Data Studio- LOAD option

304 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

򐂰 The Control Center

Control Center can be used to load data visually.
򐂰 ADMIN_CMD stored procedure
The ADMIN_CMD can be used by applications to run administrative calls and
can be used with the load command.
򐂰 An application programming interface (API) - db2load
DB2 provides an API for loading data from files. This provides an option to do
a load programmatically by using db2ApiDf.h and a method db2load.

9.5 High availability

High availability (HA) is a business requirement on the system to provide
services at all times and survive disaster, system crash, and glitches without or
with minimal interruption of the service. This is achieved by a number of
techniques such as online management of the database system, advanced
instance management, suspended I/O, and clustered servers. One of the
advanced features used for HA is failover, wherein a workload is transferred from
one system to another quickly and automatically in case of failure.

IBM provides software-based HA solutions such as DB2 HADR, WebSphere

Replication Server Q Replication, SQL Replication, and HA clustering software -
HACMP™ and Tivoli® System Automation (TSA).

The DB2 High Availability Disaster Recovery (HADR) feature is a database

replication service that uses a TCP/IP network to propagate database changes
(Data Definition Language and Data Manipulation Language statements) from a
primary to a standby database, which is monitored by a heartbeat. This HA
solution is targeted towards database systems with 24x7 availability and failover
times in the matter of seconds.

With HADR enabled, a second database copy on a different server is constantly

being updated while transactions are occurring. Transactions are logged to
ensure both database copies are protected from data loss caused by partial or
complete site failures. In a case of an outage, the standby database on the
redundant server can quickly take over. Database clients can automatically and
transparently to the application be shifted over using the alternate client re-route
feature built into the IBM DB2 Data Server client. This solution is very cost
effective and puts no special requirements on hardware. Shared storages for
example is not required, since the network is used to replicate all changed data
from one database to another. As a basic requirement, you only need exactly two
TCP/IP network-connected servers. For enhanced redundancy, multiple network

Chapter 9. Database administration 305

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

adapters can be used. As of DB2 9.7 the standby server can also be used for
read-only transactions, therefore utilizing all of the hardware at all times.

For the automatic failover the DB2-integrated component called Tivoli System
Automation for Multiplatforms (TSA MP) comes into play (often referred to as
cluster manager). This component monitors all recourses involved in both
systems - i.e. DB2 instances, databases, network interfaces, and others - and
initiates the database takeover. At the same time DB2 is aware of TSA MP and is
letting it know when a planned outage is wanted, i.e. the database was manually
stopped by an administrator. This way a database administrator can, for the most
part, use the usual DB2 commands without having to deal with the cluster
manager. Every DB2 edition supporting HADR includes all necessary software
packages as well as licenses for TSA MP and offers to install these during the
DB2 setup. It also performs the configuration of TSA MP.

Figure 9-16 illustrates a example high availability disaster recover scenario.

Figure 9-16 High Availability Disaster Recovery scenario

For detailed for more information and implementation steps, check out High
Availability and Disaster Recovery Options for DB2 on Linux, UNIX, and
Windows, SG24-7363

9.6 Autonomics
Automated task management allows automation of database management jobs
by scheduling activities according to specific requirements. This can be really

306 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

useful for performing regular maintenance tasks such as backup, space

monitoring, error checking, maintenance, etc.

IBM is striving for autonomic computing and developing highly intelligent

database systems. The DB2 autonomic computing environment is
self-configuring, self-healing, self-optimizing and self-protecting. By sensing and
responding to situations that occur, autonomic computing shifts the burden of
managing a computing environment from database administrator to technology.
Autonomic computing provides users with improved resiliency, higher ROI
(return of investment) and lower TCO (Total cost of ownership) by accelerating
the implementation of new capabilities to gain the highest value possible.

Automatic features provide a high-level summary of the capabilities that

comprise the DB2 autonomic computing environment; the following provides a
more detailed, categorized overview of the product's autonomic capabilities:

Self-Tuning Memory Manager (STMM)

The self-tuning memory manager(STMM) feature simplifies the task of memory
configuration. This feature responds to significant changes in workload by
automatically and iteratively adjusting the values of several memory
configuration parameters, including the sizes of the buffer pools, to optimize
performance. The memory tuner dynamically distributes available memory
resources among several memory consumers, including the sort function, the
package cache, the lock list, and buffer pools. By default, STTM is turned on.
You can disable STMM after creating a database by setting the database
configuration parameter self_tuning_mem to OFF.

STMM tunes a database in two different modes, DATABASE_MEMORY tuning

and no DATABASE_MEMORY tuning. With DATABASE_MEMORY tuning
memory is taken and given back to the OS as necessary, meaning the total
amount of memory allocated to a DB2 database can grow over time. Without
DATABASE_MEMORY tuning the overall amount of memory for a given
database doesn't change. Memory tuning still occurs but only between different
parameters within the database memory model.

There is a master switch to turn on STMM, to turn this on set the database
configuration parameter self_tuning_mem to ON (default for newly created
databases). Then you can choose which buffer pools and memory-related
database parameters to tune. The following memory-related database
configuration parameters can be automatically tuned:
򐂰 database_memory - Database shared memory size
򐂰 locklist - Maximum storage for lock list
򐂰 maxlocks - Maximum percent of lock list before escalation
򐂰 pckcachesz - Package cache size
򐂰 sheapthres_shr - Sort heap threshold for shared sorts

Chapter 9. Database administration 307

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 sortheap - Sort heap size

Example 9-6 on page 308 shows how to enable STMM and configure each
parameter to be managed by the STMM.

Example 9-6 Self-tuning memory manager configuration

db2inst1@db2server:~ > db2 UPDATE DB CFG FOR DATABASE invent USING
SELF_TUNING_MEM ON
db2inst1@db2server:~ > db2 ALTER BUFFERPOOL bp32 SIZE AUTOMATIC
db2inst1@db2server:~ > db2 UPDATE DB CFG FOR DATABASE invent USING locklist
AUTOMATIC

Once the SELF_TUNING_MEM parameter is set to ON, DB2 begins managing

and tuning all buffer pools and memory-related database elements. It does this
by sensing the underlying workload on the system and adjusts the memory
accordingly. STMM can adjust allocated memory from 6-120 times an hour,
depending on the changes and current settings. DB2 performs adjustments in a
clever way, allowing parameters to have a max growth of 50% or reduction of
20%. All changes are persistent and logged in the db2diag.log and STMM.log.

There are a few scenarios where STMM shine:

򐂰 Unknown memory requirements
STMM can tune a database to optimal configuration in less than an hour.
򐂰 Tuning buffer pools of multiple page sizes
STMM works with buffer pools of multiple page sizes and can easily trade
memory between the buffer pools.
򐂰 Memory varied workload
STMM constantly re-evaluates the memory requirements and will optimize
the memory usage based on the currently running workload.

Automatic Storage
The automatic storage feature simplifies storage management for table spaces.
When you create an automatic storage database, you specify the storage paths
where the database manager will place your table space data. Then, the
database manager manages the container and space allocation for the table
spaces as you create and populate them. By default automatic storage is turned
on.
򐂰 Automatic storage databases
Automatic storage is intended to make storage management easier. Rather
than managing storage at the table space level using explicit container
definitions, storage is managed at the database level and the responsibility of

308 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

creating, extending and adding containers is taken over by the database

manager.
To create a new database or alter an existing database to use automatic
storage execute the following DB2 commands.
CREATE DATABASE <dbname> AUTOMATIC STORAGE YES ON <storagePaths>
ALTER DATABASE <dbname> ADD STORAGE ON <storagePaths>
򐂰 Automatic storage table spaces
With automatic storage table spaces, storage is managed automatically. The
database manager creates and extends containers as needed up the limits
imposed by the storage paths associated with the database.
To create a new table space or alter an existing table space to use automatic
storage execute the following DB2 commands.
ALTER TABLESPACE <tablespacename> MANAGED BY AUTOMATIC STORAGE
CREATE TABLESPACE <tablespacename> MANAGED BY AUTOMATIC STORAGE

Automatic Maintenance
The database manager provides automatic maintenance capabilities for
performing database backups, keeping statistics current and reorganizing
tables and indexes as necessary. Performing maintenance activities on your
databases is essential in ensuring that they are optimized for performance
and recoverability.

Maintenance of your database may include some, or all, of the following

activities:
򐂰 Backups
When backing up a database, the database manager takes a copy of the data
in the database and stores it on a different medium in case of failure or
damage to the original. Automatic database backups helps to ensure that
your database is backed up properly and regularly so that you don't have to
worry about when to back up or know the syntax of the BACKUP command.
򐂰 Data defragmentation (table or index reorganization)
This maintenance activity can increase the efficiency with which the database
manager accesses your tables. Automatic reorganization manages an offline
table and index reorganization so that you don't need to worry about when
and how to reorganize your data.
򐂰 Data access optimization (statistics collection)
The database manager updates the system catalog statistics on the data in a
table, the data in indexes, or the data in both a table and its indexes. The
optimizer uses these statistics to determine which path is necessary to
access the data. Automatic statistics collection attempts to improve the

Chapter 9. Database administration 309

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

performance of the database by maintaining up-to-date table statistics. The

goal is to allow the optimizer to choose an access plan based on accurate
statistics.
򐂰 Statistics profiling
Automatic statistics profiling advises when and how to collect table statistics
by detecting outdated, missing, or incorrect statistics, and by generating
statistical profiles based on query feedback.

It may be time-consuming to determine whether and when to run maintenance

activities, which makes the automatic maintenance very convenient. You can
manage the enablement of the automatic maintenance features simply and
flexibly by using the automatic maintenance database configuration parameters.

The easiest way to setup automatic maintenance is to use the graphical

administrative tools. You can access the Configure Automatic Maintenance
wizard from either the Data Studio or the Control Center. Configuring your
maintenance objectives using the Data Studio is shown in Figure 9-7 on
page 291.The database manager uses these objectives to determine whether
the maintenance activities need to be done and runs only the required ones
during the next available maintenance window (a time period that you define in
which the server will not be as heavily used).

Configuration Advisor
You can use the Configuration Advisor to obtain recommendations for values of
the buffer pool size, database configuration parameters, and database manager
configuration parameters.

To use the Configuration Advisor, specify the AUTOCONFIGURE command for

an existing database, or specify AUTOCONFIGURE as an option of the
CREATE DATABASE command.

You can display the recommended values or apply them by using the APPLY
option of the CREATE DATABASE command. The recommendations are based
on input that you provide and system information that the advisor gathers.

When you create a database, this tool is automatically run to determine and set
the database configuration parameters and the size of the default buffer pool
(IBMDEFAULTBP). The values are selected based on system resources and the
intended use of the system. This initial automatic tuning means that your
database performs better than an equivalent database that you could create with
the default values. It also means that you will spend less time tuning your system
after creating the database.

310 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

You can run the Configuration Advisor at any time (even after your databases
are populated) to have the tool recommend and optionally apply a set of
configuration parameters to optimize performance based on the current system
characteristics. You can use the graphical database administration tools to run
the Configuration Advisor. Figure 9-4 on page 287 shows the Data Studio
Configuration Advisor Wizard

The values suggested by the Configuration Advisor are relevant for only one
database per instance. If you want to use this advisor on more than one
database, each database must belong to a separate instance.

Data compression
Both tables and indexes can be compressed to save storage. Compression is
fully automatic; once you specify that a table or index should be compressed
using the COMPRESS YES clause of the CREATE TABLE, ALTER TABLE,
CREATE INDEX or ALTER INDEX statements, there is nothing more you must
do to manage compression. Temporary tables are compressed automatically;
indexes for compressed tables are also compressed automatically, by default.
Data Compression is discussed in further in 11.2, “Data compression” on
page 383.

Monitoring database health

The health monitor is a server-side tool that proactively monitors situations or
changes in your database environment that could result in performance
degradation or a potential outage. A range of health information is presented
without any form of active monitoring on your part. If a health risk is encountered,
the database manager informs and advises you on how to proceed. The health
monitor gathers information about the system by using the snapshot monitor and
does not impose a performance penalty. Further, it does not turn on any
snapshot monitor switches to gather information.

Utility throttling
This feature regulates the performance impact of maintenance utilities so that
they can run concurrently during production periods. Although the impact policy
for throttled utilities is defined by default, you must set the impact priority if you
want to run a throttled utility. The throttling system ensures that the throttled
utilities run as frequently as possible without violating the impact policy.
Currently, you can throttle statistics collection, backup operations, rebalancing
operations, and asynchronous index cleanup.

Chapter 9. Database administration 311

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

9.7 Workload management

More and more systems get consolidated due to the increasing calculating power
of CPUs. Although hardware utilization is increased and costs are reduced,
management complexity is increased requiring more knowledge and efficient
handling for workloads running on the system. When consolidating databases
from systems holding different types of data and accessed by differently
behaving workloads, a sophisticated workload management solution is required
to ensure a stable and predictable execution environment.

As of DB2 Data Server 9.5 Workload Management (WLM) is built right into the
DB2 engine allowing administrators to monitor and control database activities,
such as DDL and DML statements over their full life cycle. Through definable
rules the engine can be programmed to automatically filter certain workloads and
apply execution priorities to control concurrency or activity through setup of
thresholds. WLM can explicitly control CPU usage among executing work, detect
and prevent so-called “run away” queries, which for example are exceeding the
predicted or configured number of rows returned, the execution time or the
estimated execution costs. Specifically on the AIX platform DB2 workload
management does not stop on a database level and can be tightly integrated with
the operating system's workload management.

For more details and setup information refer to the IBM DB2 9.7 for Linux, UNIX,
Windows, Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

9.8 Database management tools

In this section we introduce the graphical tools available for managing DB2
database servers. Although you can do all these tasks using the command
prompt, these tools play an important role in allowing administrator to manage
the database graphically.

DB2 supports impressive suite of tools for database management and services.
All the daily database operations can be easily and effectively done using
following GUI tools and wizards:
򐂰 Control Center
򐂰 Optim Development Studio
򐂰 Optim Database Administrator
򐂰 IBM Data Studio
򐂰 Data Studio Administration Console

312 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

9.8.1 DB2 Control Center

IBM DB2 Control Center is the central point from which you can manage your
family of DB2 databases, running on an array of operating systems in your
workplace. A user-friendly graphical interface makes your job easier by guiding
you through each of the steps in managing DB2 Database.

DB2 Control Center provides a common interface for managing DB2 databases
on different platforms. You can run DB2 commands, create DDL statements, and
execute DB2 utilities. DB2 Control Center provides point-and-click navigation
capabilities make it easy to find objects, whether you have hundreds or tens of
thousands in your database environment. It can be used to administer system,
instances, tables, views, indexes, triggers, user-defined types, user-defined
functions, packages, aliases, users, or groups.

It is tightly coupled with other DB2 tools; Figure 9-17 shows a hierarchy of
database objects on the left hand panel, and details on right hand panel. The
Control Center can started by calling the db2cc command.

Figure 9-17 DB2 Control Center

The following tools can be launched from the Control Center Tools menu:
򐂰 Replication Center
򐂰 Satellite Administration Center Command Center
򐂰 Command Editor
򐂰 Task Center
򐂰 Health Center

Chapter 9. Database administration 313

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 Journal
򐂰 License Center
򐂰 Configuration Assistant

DB2 Control Center also provides set of wizards for completing specific
administration tasks by taking you through each step one at a time. The following
DB2 wizards are available though the Control Center:
򐂰 Add database partitions launch pad
򐂰 Backup wizard
򐂰 Create database wizard
򐂰 Create database with automatic maintenance
򐂰 Create table space wizard
򐂰 Create table wizard
򐂰 Design Advisor
򐂰 Load wizard
򐂰 Configuration Advisor
򐂰 Restore data wizard
򐂰 Configure database logging wizard
򐂰 Set up activity monitor wizard
򐂰 Set up high availability disaster recovery databases
򐂰 Configure automatic maintenance

9.8.2 IBM Optim and Data Studio Suite Overview

IBM Optim and Data Studio provides a new comprehensive suite of integrated
tools for development database administrators and application developers. The
tooling for each of these roles is essentially divided between two main views in
the Data Perspective which are the Data Project Explorer, and the Data Source
Explorer, shown in Figure 9-18 on page 315. There are also features to help
team members share resources.

A benefit to these new tools is that you can have one single environment to
develop and manage your database, which required separate tools in the past.
This tool also supports multiple database servers, such as DB2 for LUW, i5/zOS,
Apache Derby, Informix IDS and others, this make managing multiple database
much easier.

The IBM Optim and Data Studio tool suite is built on the Eclipse platform and, as
such, is said to be an Eclipse-based development environment. The Eclipse
platform is a framework that allows integrated development environments (IDE to
be created; plug-ins exist to allow development in Java, C/C++, PHP, COBOL,
Ruby, and more.

314 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

Figure 9-18 IBM Optim and Data Studio tool suite

There are two paid versions, the Optim Database Administrator and the Optim
Development Studio, and one free version of IBM Data Studio.

Optim Database Administrator

The Optim Database Administrator, formerly called Data Studio Administrator, is
another GUI tool under the new suite of database management GUI tools. This
tool is a powerful and flexible tool that helps you manage your database objects.
The Optim Database Administrator simplifies DB2 database administration by
providing an easy-to-use interface and built-in analysis to automate complex
database changes and conversions with confidence. Its capabilities make it easy
to perform everyday tasks such as object changes, utilities, commands,
configuration changes, and granting or revoking of security privileges and
enables immediate script execution or the ability to save the script for later
execution.

You can use Optim Database Administrator to perform common database

administration tasks within the Eclipse environment, such as:
򐂰 Starting or stopping a DB2 instance
򐂰 Backing up a database
򐂰 Creating database objects
򐂰 Modifying database objects
򐂰 Changing configuration settings
򐂰 Granting and revoking security privileges

Chapter 9. Database administration 315

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

򐂰 Running DB2 commands and utilities

Optim Database Administrator provides an integrated end-to-end solution that

enables you to seamlessly work with multiple products simultaneously.

Optim Database Administrator 2.2 is on Eclipse 3.4, and here are some of the
products it can shell share with:
򐂰 Optim Development Studio 2.2
򐂰 Optim Query Tuner 2.2
򐂰 InfoSphere Data Architect 7.5.x.x

Optim Development Studio

The Optim Development Studio, formerly called Data Studio Developer, is part of
the new suite of GUI tools for managing DB2 data and database application. The
Optim Development Studio enables database developers to rapidly develop high
quality code by providing a feature rich database development environment. This
environment simplifies database development by providing procedure building
and debugging features along with database object management, project
management, and optimization tools with Optim pureQuery Runtime.

For data application developers, Optim Development Studio provides the

following key features. Working in a data development project in the Data Project
Explorer, you can:
򐂰 Develop pureQuery applications in a Java project
򐂰 Use wizards and editors to create, test, debug, and deploy routines, such as
stored procedures and user-defined functions
򐂰 Use the SQL builder and the SQL editor to create, edit, and run SQL queries
򐂰 Use Visual Explain to tune routines and SQL queries
򐂰 Use the Routine debugger to debug stored procedures
򐂰 Create Web services that expose database operations (SQL SELECT and
DML statements, XQuery expressions, or calls to stored procedures) to client
applications
򐂰 Use wizards and editors to develop XML applications
򐂰 Develop SQLJ applications in a Java project

The Optim Development Studio provides the following key features for database
object management. Typically these tasks are done on test databases that you
are using to test your applications.

Working in the Data Source Explorer, you can:

򐂰 Connect to data sources and browse data objects and their properties

316 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

򐂰 Use editors and wizards to create and alter data objects

򐂰 Modify privileges for data objects and authorization IDs
򐂰 Copy database objects
򐂰 Drop data objects from databases
򐂰 Analyze the impact of your changes
򐂰 Work with data, including extracting and loading data and inserting XML data
into XML columns
򐂰 Use data diagrams to visualize the relationships between data objects

If you are working on a large team, you can use the following features to enable
team members to share resources:
򐂰 You can share data development projects using supported source code
control systems.
򐂰 You can share database connection information by importing and exporting
this information to XML files.
򐂰 You can customize the user interface to enable and disable visible controls
and defaults.

Optim Development Studio 2.2 is on Eclipse Version 3.4, and here are some of
the products it can shell share with:
򐂰 Rational Application Developer for WebSphere Software 7.5.x.x
򐂰 Optim Database Administrator 2.2
򐂰 InfoSphere Data Architect 7.5.x.x

IBM Data Studio

This is a freely downloadable version called IBM Data Studio, provides an
integrated development environment for both administrating and developing with
IBM data servers. The IBM Data Studio is available in two packaging options:
򐂰 The integrated development environment (IDE) package includes all
administrative capabilities as well as an integrated Eclipse development
environment for Java, XML, and Web services.
򐂰 The new stand-alone package is a lighter weight offering designed
specifically for administrators to get up and running quickly and easily. This
package does not include some of the data development capabilities
available in the IDE package and cannot be installed with other products to
extend its capabilities.

For more information and software download visit:

http://www.ibm.com/software/data/optim/

Chapter 9. Database administration 317

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

IBM Data Studio administration console

The IBM Data Studio Administration Console provides a no charge rich Web
interface for monitoring your DB2 servers. Use its database health and
availability monitoring features to get a summary of system health, view
dashboard matrixes, investigate alerts, and troubleshoot problems using expert
recommendations. The Data Server Administration Console monitors the health
and availability of your connected DB2 data server databases on Linux, UNIX,
Windows, and z/OS systems.

With the Data Studio Administration Console you can:

򐂰 View system health at a glance.
– Connect to and monitor multiple databases across different platforms from
a single console.
– Configure thresholds for warnings and alerts for key performance
indicators.
򐂰 Drill down into alerts and warnings for a database.
– Examine the details when problems occur with key performance
indicators.
– Filter the display of statistics by time period, database, or application.
򐂰 Browse alert history.
– Collect and retain alert history for 72 hours.
– Filter the display of statistics by time period, database, or alert type.
򐂰 Use expert recommendations for problem solving.
– Review expert recommendations to help solve the problems that are
causing a specific alert.
– View the system and database parameters and statistics associated with a
warning or alert.

You can also use the Data Studio Administration Console to monitor Q
replication and event publishing, generate replication health reports, and perform
basic replication operations.

You can launch Data Studio Administration Console from the IBM Data Studio
Developer user interface so that you can monitor IBM data servers for status
including database availability, dashboards, and alerts. Figure 9-19 on page 319
shows the Data Studio Administration Console dash.

318 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ch09.fm

Figure 9-19 Data Studio Administration Console - Dashboard view

Chapter 9. Database administration 319

7093ch09.fm Draft Document for Review October 16, 2009 10:22 am

320 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

10

Chapter 10. Testing and tuning

In this chapter we discuss the steps to verify that data and application
functionality were ported completely and correctly. This includes the following:
򐂰 Test planning
򐂰 Data checking
򐂰 Code and application testing
򐂰 Troubleshooting

We also provide information on how you can check that system behavior has not
changed in an undesired way.

Further, methods and tools available for DB2 to tune the database in order to
achieve optimal performance are discussed.

© Copyright IBM Corp. 2009. All rights reserved. 321

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

10.1 Test planning

The test planning stages detail the activities, dependencies, and efforts required
to conduct the test of the converted solution.

10.1.1 Principles of software tests

Keep in mind the following principles of software tests in general:
򐂰 It is not possible to test a non-trivial system completely.
򐂰 Tests are optimizing processes regarding completeness.
򐂰 Always test against expectations.
򐂰 Each test must have reachable goals.
򐂰 Test cases have to contain reachable and non-reachable data.
򐂰 Test cases must be repeatable.
򐂰 Test cases have to be archived in the configuration management system as
well as source code and documentation.

10.1.2 Test documentation

The test documentation is a very important part of the project. The ANSI/IEEE
Standard 829-1998 for Software Test Documentation describes the exact
content of the test document. Refer to this document for more details.

10.1.3 Test phases

Series of well-designed tests should validate all stages of the conversion
process. A detailed test plan should describe all the test phases, scope of the
tests, validation criteria, and specify the time frame. To ensure that the
applications operate in the same manner as they did in the source database, the
test plan should include data conversion, functional and performance tests, as
well as the following other post conversion assessments:
򐂰 Data conversion testing
򐂰 Functional testing
򐂰 Integration testing
򐂰 Performance testing
򐂰 Volume/load stress testing
򐂰 Acceptance testing
򐂰 Post conversion tests

322 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

10.1.4 1 Time planning and time exposure

The time planning should be based on realistic and validated estimates. If the
estimates for the conversion of the application and database are inaccurate, the
entire project plan will slip.

It is always best to tie all test dates directly to their related conversion activity
dates. This prevents the test team from being perceived as the cause of a delay.
For example, if system testing is to begin after delivery of the final build, then
system testing begins the day after delivery. If the delivery is late, system testing
starts from the day of delivery, not on a specific date. This is called dependent or
relative dating.

Figure 10-1 shows the test phases during a typical conversion project. The
definition of the test plans happen at a very early moment. The test cases, and all
subsequent tasks, must be done for all test phases.

Create Test Plan

Create Test Cases

Prepare Test Items

Prepare Infrastructure

Test Run in Test Run in

MySQL Environment DB2 Environment

Compare Test Results

Figure 10-1 Test phases during a conversion project

The time exposure of tests depends on the availability of an existing test plan
and already prepared test items. The efforts depend also on the degree of
changes during the application and database conversion.

Note: Test efforts can range between 50% and 70% of the total conversion
effort.

Chapter 10. Testing and tuning 323

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

10.2 Data checking techniques

Data movement is the first focus for any conversion project. Without having all
your tables and data properly moved over, all other conversion testing is in vain.

The testing process should detect if all rows were imported into the target
database, ensure that all data type conversions were successful, and check
random data byte-by-byte. The data checking process should be automated by
appropriate scripts. When testing data conversion results you should:
򐂰 Check IMPORT/LOAD messages for errors and warnings.
򐂰 Count the number of rows in source and target databases and compare them.
򐂰 Prepare scripts that perform data checks.
򐂰 Involve data administration staff familiar with the application and its data to
perform random checks.

10.2.1 IMPORT/LOAD messages

You should always check the messages generated by the IMPORT or LOAD
commands. Example 10-1 presents messages generated by the sample IMPORT
command. You should read not only the summary at the end of the listing, but
also pay attention to the warning messages.

Example 10-1 Sample IMPORT message

db2inst1@db2server:~> db2 IMPORT FROM table01Data.txt OF DEL REPLACE INTO
table01

SQL3109N The utility is beginning to load data from file "table01Data.txt".

SQL3148W A row from the input file was not inserted into the table. SQLCODE "-545" was
returned.

SQL0545N The requested operation is not allowed because a row does not
satisfy the check constraint "DB2INST1.TABLE01.SQL090915100543100".
SQLSTATE=23513

SQL3185W The previous error occurred while processing data from row "2" of the input
file.

SQL3117W The field value in row "3" and column "1" cannot be converted to a SMALLINT
value. A null was loaded.

SQL3125W The character data in row "4" and column "2" was truncated because the data is
longer than the target database column.

SQL3110N The utility has completed processing. "4" rows were read from the input file.

324 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

SQL3221W ...Begin COMMIT WORK. Input Record Count = "4".

SQL3222W ...COMMIT of any database changes was successful.

SQL3149N "4" rows were processed from the input file. "3" rows were
successfully inserted into the table. "1" rows were rejected.

Number of rows read = 4

Number of rows skipped = 0
Number of rows inserted = 3
Number of rows updated = 0
Number of rows rejected = 1
Number of rows committed = 4

As shown in the summary, during the import process one record from the input
file was rejected, and three were inserted into the database. To understand the
nature of the warnings, you should look into the data source file and the table
definition (db2look command). For Example 10-1 on page 324, the table
definition is presented in Example 10-2, and the data file in Example 10-3

Example 10-2 Table definition for Example 10-1

db2> CREATE TABLE TABLE01 (
C1 SMALLINT,
C2 CHAR(3),
C3 SMALLINT CHECK( C3 IN (1,2,3)))

Example 10-3 Data file for Example 10-1

1,"abc",1
2,"abc",4
32768,"abc",2
4,"abcd",3

The first row from the input file (Example 10-3) was inserted without any
warnings. The second row was rejected because it violated check constraints
(warnings SQL3148W, SQL0545N, SQL3185W). A value of 32768 from the third
row was changed to null because it was out of SMALLINT data type range
(warning SQL3117W) and string abcd from the last row was truncated to abc
because it was longer than the relevant column definition (warning SQL3125W).

The LOAD utility generates messages in a similar format, but because it is

designed for speed, it bypasses the SQL engine, and inserts data directly into
table spaces without constraint checking. Inserting the same table01.unl file
(Example 10-3) into table01 (Example 10-2) with the LOAD utility generates

Chapter 10. Testing and tuning 325

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

messages without SQL3148W, SQL0545N, SQL3185W warnings as shown in

Example 10-4.

Example 10-4 LOAD messages

db2inst1@db2server:~> db2 LOAD FROM table01Data.txt OF DEL REPLACE INTO table02

SQL3109N The utility is beginning to load data from file

"/home/db2inst1/DB2Scripts/chp10/table01Data.txt".

SQL3500W The utility is beginning the "LOAD" phase at time "09/15/2009

10:15:23.814793".

SQL3519W Begin Load Consistency Point. Input record count = "0".

SQL3520W Load Consistency Point was successful.

SQL3117W The field value in row "F0-3" and column "1" cannot be converted to a SMALLINT
value. A null was loaded.

SQL3125W The character data in row "F0-4" and column "2" was truncated
because the data is longer than the target database column.

SQL3227W Record token "F0-3" refers to user record number "3".

SQL3227W Record token "F0-4" refers to user record number "4".

SQL3110N The utility has completed processing. "4" rows were read from the
input file.

SQL3519W Begin Load Consistency Point. Input record count = "4".

SQL3520W Load Consistency Point was successful.

SQL3515W The utility has finished the "LOAD" phase at time "09/15/2009
10:15:23.836019".

SQL3107W There is at least one warning message in the message file.

Number of rows read = 4

Number of rows skipped = 0
Number of rows loaded = 4
Number of rows rejected = 0
Number of rows deleted = 0
Number of rows committed = 4

A table that has been created with constraints is left by the LOAD command in
check pending state. Accessing the table with SQL queries generates a warning:
SQL0668N Operation not allowed for reason code "1" on table <TABLE_NAME>.
SQLSTATE=57016.

326 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

The SET INTEGRITY SQL statement should be used to move loaded tables into
a usable state. Example 10-5 shows a way to validate constraints. All rows that
violated constraints will be moved to exception table table01_e.

Example 10-5 Turning integrity checking on

db2inst1@db2server:~> db2 CREATE TABLE table02_exp LIKE table02
db2inst1@db2server:~> db2 SET INTEGRITY FOR table02 IMMEDIATE CHECKED FOR
EXCEPTION IN table02 USE table02_exp
SQL3602W Check data processing found constraint violations and moved them to
exception tables. SQLSTATE=01603

The SET INTEGRITY statement has many options like turning integrity on only
for new data, turning integrity off, or specifying exception tables with additional
diagnostic information. To read more about the SET INTEGRITY command refer
to:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

IBM Data Movement Tool LOAD messages and the SET

INTEGRITY script
When using the IBM Data Movement Tool to convert, it is simple to review the
data conversion logs. The IBM Data Movement Tool uses the DB2 LOAD
command to load the converted data into DB2. All load output logs are located in
the msg directory in the conversion project output directory. There is a file for
each table that was loaded and the log output is in the same format as described
in Example 10-4 on page 326.

The IBM Data Movement tool also generates a script called db2checkpending.sql
in the conversion project output directory, which when executed runs a SET
INTEGRITY on each of the tables.

10.2.2 Data checking

Scripts performing logical data integrity checks automate the data verification
process and save administration efforts. For small tables (with less that 50,000
rows) you can write a program that compares data byte-by-byte. The program
can extract sorted rows from MySQL and DB2 to files in the same ASCII format.
The files should be then binary compared (on Linux use the diff command) and
checked to determine if they are the same. For larger tables, comparing all rows
byte-by-byte can be very inefficient. The data conversion should be evaluated by
comparing aggregate values like the number of rows. To do this you can create a
special table for storing the information about the number of rows in the source
MySQL database. Table ck_row_count presented in Example 10-6 on page 328
can be used for that purpose.

Chapter 10. Testing and tuning 327

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Example 10-6 Table for storing number of rows (MySQL)

CREATE TABLE ck_row_count (
tab_name VARCHAR(30), -- table name
row_count INT, -- number of rows
sys_name CHAR(3), -- code to distinguish the system: MYS or DB2
time_ins DATE -- time when the count was performed
)

For each table you should count the number of rows and store the information in
the CK_ROW_COUNT table. The following INSERT statement can be used for
that purpose:
INSERT INTO ck_row_count SELECT 'tab_name', COUNT(*), 'MYS', sysdate() FROM
tab_name

The table ck_row_counts and its data can be manually converted to the target
DB2 database. Example 10-7presents the DB2 version of the table.

Example 10-7 Table for storing number of rows (DB2)

CREATE TABLE ck_row_count (
tab_name VARCHAR(30),
row_count INT,
sys_name CHAR(3),
time_ins TIMESTAMP
)

On the DB2 system, you should repeat the counting process with the equivalent
INSERT statement:
INSERT INTO ck_row_count SELECT 'tab_name', COUNT(*), 'DB2', CURRENT
TIMESTAMP FROM tab_name

After performing the described steps, DB2 table CK_ROW_COUNT should

contain information about the number of rows counted in MySQL and DB2
databases. The records in the table should look similar to Example 10-8.

Example 10-8 Sample table ck_row_counts contents

SELECT * FROM ck_row_count

TAB_NAME ROW_COUNT SYS_NAME TIME_INS

------------- ---------- -------- --------------------------
inventory 703 DB2 2009-09-15-10.53.46.573998
services 808 DB2 2009-09-15-10.55.18.782980
groups 6 DB2 2009-09-15-10.55.45.852181
owners 502 DB2 2009-09-15-10.56.02.341948
locations 140 DB2 2009-09-15-10.56.15.269177
severity 5 DB2 2009-09-15-10.56.31.337119

328 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

status 7 DB2 2009-09-15-10.56.51.599962

Having the information about the number of rows in a SQL table is convenient
because with a single query you can get the table names that contain a different
number of rows in the source and target database:
SELECT tab_name FROM (SELECT DISTINCT tab_name, row_count FROM
ck_row_count) AS t_temp GROUP BY t_temp.tab_name HAVING(COUNT(*) > 1)

The presented approach for comparing the number of rows can be extended for
additional checking such as comparing the sum of numeric columns. Here are
the steps that summarize the technique:
1. Define check sum tables on the source database and characterize scope of
the computation.
2. Perform the computation and store the results in the appropriate check sum
tables.
3. Convert the check sum tables as other user tables.
4. Perform equivalent computations on the target system, and store the
information in the converted check sum tables.
5. Compare the computed values.

Table 10-1 provides computations for selected database types. The argument for
the DB2 SUM() function is converted to DECIMAL type. This occurs because in
most cases the SUM() function returns the same data type as its argument,
which can cause arithmetic overflow. For example, when calculating the sum on
an INTEGER column, if the result exceeds the INTEGER data type range, error
SQL0802N is generated: Arithmetic overflow or other arithmetic exception
occurred. Converting the argument to DECIMAL eliminates the error.

Table 10-1 Aggregations for data conversion verification

Data type MySQL operation DB2 operation

numeric(<precision>,<scale>) sum(val) sum(cast(val as

decimal(31,<scale>)))

date sum(trunc( val - to_date sum(cast(days(val) as

('0001/01/02','yyyy/mm/dd'))) decimal(31,1)))

variable length character sum(length(val)) sum(cast(length(val) as

decimal(31,0)))

fixed length character sum(length(rtrim(val))) sum(cast(length(rtrim(val))as

decimal(31,0)))

Chapter 10. Testing and tuning 329

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Data checking with IBM Data Movement Tool

The IBM Data Movement Tool can simplify the process of data checking your
converted data. The tool generates a script called rowcount in the conversion
project output directory. When executed, this script connects to the MySQL and
DB2 databases and counts the rows for each table in each database. The output
is logged into a file <database_name>.tables.rowcount for easy comparison.
Example 10-9 shows the output file for our sample conversion.

Example 10-9 Sample output inventory.tables.rowcount

db2inst1@db2server:/opt/ibm/IBMDataMovementTool/migr> cat
inventory.tables.rowcount
mysql db2
mysql.groups : 6 "ADMIN"."GROUPS" : 6
mysql.inventory : 703 "ADMIN"."INVENTORY" : 703
mysql.locations : 140 "ADMIN"."LOCATIONS" : 140
mysql.owners : 502 "ADMIN"."OWNERS" : 502
mysql.services : 808 "ADMIN"."SERVICES" : 808
mysql.severity : 5 "ADMIN"."SEVERITY" : 5
mysql.status : 7 "ADMIN"."STATUS" :
7

10.3 Code and application testing

The most important part of the testing process is to verify that each component of
the system functions as it did prior to conversion. All components of the RDBMS
system should be verified for their functionality.

10.3.1 Application code check

The scope of application testing depends on the converted application. For
self-built applications, the testing process should be started with the application
queries. All queries should be independently tested to ensure that they return the
expected results. With the application queries successfully converted, the
surrounding client programs should be rebuilt, and the application should be
tested against the target database. Each module of the application and possibly
each screen form should be run and be checked for errors or improper
functionality. All supported interfaces should also be checked.

An important issue is documentation of all the test conditions. These include

operations performed, application screens opened, input data used for testing,
and results. For larger projects, the documenting component may become
overwhelming increasing the need for specialized software. As mentioned
earlier, by definition, new applications cannot be fully tested. In the conversion

330 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

project, the application testing is an iterative process of planning, designing test

cases, executing the test cases, and finally evaluating and analyzing the results.

Along with the functional testing, the application should also be checked against
performance requirements. Since there are many architectural differences
between MySQL and DB2, some SQL operations might require further
optimization. Observing the performance differences in the early testing stages
increases the chance of preparing more optimal code for the new environment.

Before going into production, the converted database should be verified under
high volumes and loads. These tests should emulate the production
environment, and can determine if further application or database tuning is
necessary. The stress load can also reveal other hidden problems such as
locking issues, which can be observed only in a production environment.

10.3.2 Security testing

Before going into production, security must be verified. MySQL handles security
differently than DB2, so it is not trivial to compare the user rights between the two
systems.

MySQL users and privileges are resolved in DB2 with operating system users
and groups. A list of MySQL users should be compared to the equivalent DB2
operating system users. All of DB2's authorities should be verified to allow the
correct individuals to connect to the database. Privileges for all database objects
should also be verified.

10.4 Troubleshooting
The first step of problem determination is to know what information is available to
you. When DB2 performs an operation, an associated return code is returned.
The return code is displayed to the user in the form of an informational or error
message. These messages are logged into diagnostic files depending on the
diagnostic level set in the DB2 configuration. In this section we discuss the DB2
diagnostic logs, error message interpretations, and tips which may help with
problem determination, troubleshooting, and resolutions to specific problems.

The following actions should be taken when experiencing a DB2 related problem:
򐂰 Check related messages
򐂰 Explain error codes
򐂰 Check documentation
򐂰 Search through available Internet resources

Chapter 10. Testing and tuning 331

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

򐂰 Review APARs for current FixPack level

򐂰 Use available tools to narrow the problem
򐂰 Ask IBM for support

10.4.1 Interpreting DB2 informational messages

Start your investigation from the return code. DB2 provides a return code for
every operation performed in the form of CCCnnnnnS. The prefix CCC identifies
the DB2 component that is returning the message; the nnnnn is a three to five
digit error code which is also referred to as SQLCODE; and the S is a severity
indicator. For example, in SQL0289N, the SQL identifier represents a message
from the Database Manager, the SQLCODE is 0289, and N indicates an error
message.

Here is the complete list for DB2 error messages to prefix your reference:
򐂰 ASN: Replication messages
򐂰 CCA: Client Configuration Assistant messages
򐂰 CLI: Call Level Interface messages
򐂰 DB2: Command Line Processor messages
򐂰 DBA: Control Center and Database Administration Utility messages
򐂰 DBI: Installation or configuration messages
򐂰 EXP: Explain utility messages
򐂰 FLG: Information Catalog Manager messages
򐂰 LIC: DB2 license manager messages
򐂰 SAT: Satellite messages
򐂰 SPM: Synch Point Manager messages
򐂰 SQJ: Embedded SQLJ in Java messages
򐂰 SQL: Database Manager messages

The two severity indicators are:

򐂰 W: Indicates warning or informational messages
򐂰 N: Indicates error messages

DB2 also provides detailed information for each message. The full error
message describes the nature of the problem in detail along with potential user
responses. To display the full message for the DB2 return, you can use the DB2
command db2 ? error-code in Linux or AIX. Since ? (question mark) is a special
character; you must separate the DB2 command and the error code with a
double quote ("). See Example 10-10.

Example 10-10 Explaining error codes

db2 "? sql0289"
SQL0289N Unable to allocate new pages in tablespace "<tablespace-name>".

332 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Explanation:

Explanation:

One of the following conditions is true on one or more database partitions:

1 One of the containers assigned to this SMS table space has

reached the maximum file size. This is the likely cause of the
error.

2 All the containers assigned to this DMS table space are full.
This is the likely cause of the error.
[...]

You can find full information about the DB2 message format, and a listing of all
the messages in Messages Reference, Volumes 1 and 2, SC27-2450-00, and
SC27-2451-00, available online at:
http://www.ibm.com/support/docview.wss?rs=71&uid=swg27015148

10.4.2 DB2 Tools for Troubleshooting

The following tools are available to help collect, format or analyze diagnostic
data. We'll discuss some of these in more details in the next sections.
򐂰 db2dart
The db2dart command can be used to verify the architectural correctness of
databases and their associated objects. It can also be used to display the
contents of database control files in order to extract data from tables that
might otherwise be inaccessible.
򐂰 db2diag
The db2diag tool serves to filter and format the volume of information
available in the db2diag log files. Filtering records in db2diag log files can
reduce the time required to locate the records needed when troubleshooting
problems.
򐂰 db2greg
You can view and edit the Global Registry with the db2greg tool.
򐂰 db2level
The db2level command helps you determine the version and service level
(build level and fix pack number) of your DB2 instance.
򐂰 db2look

Chapter 10. Testing and tuning 333

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

There are many times when it is advantageous to be able to create a

database that is similar in structure to another database. For example, rather
than testing out new applications or recovery plans on a production system, it
makes more sense to create a test system that is similar in structure and
data, and perform tests against the test system without adversely affecting
the production system. You can use the db2look tool to extract the required
DDL statements needed to reproduce the database objects of one database
in another database. The tool can also generate the required SQL statements
to replicate statistics from one database to another, and any other statements
needed to replicate the database configuration, database manager
configuration, and registry variables.
򐂰 db2ls
With the ability to install multiple copies of DB2 products on your system with
extended flexibility to install particular DB2 products and features in any paths
of your choice, you need a tool to help you keep track of what is installed and
where it is installed. On supported Linux and UNIX operating systems, the
db2ls command lists the DB2 products and features installed on your system,
including the DB2 Version 9 HTML documentation.
򐂰 db2pd
The db2pd tool is used for troubleshooting because it can return quick and
immediate information from the DB2 memory sets.
򐂰 db2support
When it comes to collecting information for a DB2 problem, the most
important DB2 utility is db2support. The db2support utility automatically
collects all DB2 and system diagnostic information available. It also has an
optional interactive “Question and Answer” session, which poses questions
about the circumstances of your problem.
򐂰 Traces
If you experience a recurring and reproducible problem with DB2, tracing
sometimes allows you to capture additional information about it. Under
normal circumstances, you should only use a trace if asked to by IBM
Software Support. The process of taking a trace entails setting up the trace
facility, reproducing the error and collecting the data.

10.4.3 DB2 diagnostic logs

DB2 logs every return code in diagnostic logs based on the diagnostic level set in
the database manager configuration. When investigating DB2 problems, the
essential information can be found in diagnostic log files generated by DB2.
These logs are:

334 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

򐂰 Administration Notification log or Windows Event log

򐂰 db2diag.log
򐂰 Trap files
򐂰 DB2 dump files
򐂰 Core files (Linux/UNIX only)

Administration Notification log

DB2 also provides diagnosis information to the administration notification log in
case of a failure. On Linux and UNIX platforms, the administration notification log
is a text file called <instance>.nfy, where <instance> is the name of the DB2
instance. It is located in the $HOME/sqllib/db2dump directory. On Windows, all
administration notification messages are written to the Event Log.

The DBM configuration parameter NOTIFYLEVEL specifies the level of

information to be recorded:
򐂰 0 - No administration notification messages captured (not recommended)
򐂰 1 - Fatal or unrecoverable errors
򐂰 2 - Immediate action required
򐂰 3 - Important information, no immediate action required (default)
򐂰 4 - Informational messages

DB2 is not the only one that can write to the notification logs, tools such as the
Health Monitor, Capture and Apply programs, and user applications can also
write to these logs using the db2AdminMsgWrite API function.

db2diag.log
The db2diag.log is the most often used file for DB2 problem investigation. You
can find this file in the DB2 diagnostic directory, defined by the DIAGPATH
variable in the database manager configuration. If the DIAGPATH parameter is
not set, by default the directory is located at:

Linux and UNIX:

$HOME/sqllib/db2dump

where $HOME is the home directory of the DB2 instance. Windows:

<INSTALL PATH>\<DB2INSTANCE>

where <INSTALL PATH> is the directory where DB2 is installed, and

<DB2INSTANCE> is the name of the DB2 instance.

The database manager configuration parameter DIAGLEVEL controls how much

information is logged to the db2diag.log. Valid values can range from 0 to 4:
򐂰 0 - No diagnostic data captured

Chapter 10. Testing and tuning 335

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

򐂰 1 - Severe errors only

򐂰 2 - All errors
򐂰 3 - All errors and warnings (default)
򐂰 4 - All errors, warnings and informational messages

Most of the time, the default value is sufficient for problem determination. In
some cases, especially on development or test systems you can set the
parameter to 4, in order to collect all informational messages. However, ensure
to pay attention to the database activities and the size available on the file
system, since this may cause performance issues due to the large amounts of
data recorded into the file. Setting DIAGLEVEL to 4 may also make the file very
large and harder to read.

Information within this db2diag.log includes:

򐂰 A diagnostic message (beginning with DIA) explaining the reason for the error
򐂰 Application identifiers, which allow matching up error entries with
򐂰 corresponding application or DB2 server processes
򐂰 Any available supporting data such as SQLCA data structures, and pointers
to the location of any extra dump or trap files
򐂰 Administrative events, i.e. backup/restore start and finish

Example 10-11 contains an extract of a db2diag.log taken at DIAGLEVEL 3.

Example 10-11 Example of db2dial.log file

1 2007-05-18-14.20.46.973000-240 2 I27204F655 3 LEVEL: Info
4 PID : 3228 5 TID : 8796 6 PROC : db2syscs.exe
7 INSTANCE: DB2MPP 8 NODE : 002 9 DB : WIN3DB1
10 APPHDL : 0-51 11 APPID: 9.26.54.62.45837.070518182042
12 AUTHID : UDBADM
13 EDUID : 8796 14 EDUNAME: db2agntp (WIN3DB1) 2
15 FUNCTION: DB2 UDB, data management, sqldInitDBCB, probe:4820
16 DATA #1 : String, 26 bytes
Setting ADC Threshold to:
DATA #2 : unsigned integer, 8 bytes
1048576

Explanations of the db2diag.log entries are included below. The numbers bolded
in the example correspond to the following numbers:
1. A timestamp and timezone for the message.
2. The record ID field. The recordID of the db2diag log files specifies the file
offset at which the current message is being logged (for example, “27204”)
and the message length (for example, “655”) for the platform where the DB2
diagnostic log was created.

336 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

3. The diagnostic level associated with an error message. For example, Info,
Warning, Error, Severe, or Event
4. The process ID
5. The thread ID
6. The process name
7. The name of the instance generating the message.
8. For multi-partition systems, the database partition generating the message.
(In a non-partitioned database, the value is “000”.)
9. The database name
10.The application handle. This value aligns with that used in db2pd output and
lock dump files. It consists of the coordinator partition number followed by the
coordinator index number, separated by a dash.
11.Identification of the application for which the process is working. In this
example, the process generating the message is working on behalf of an
application with the ID 9.26.54.62.45837.070518182042.
TCP/IP-generated application ID is composed of three sections
i. IP address: It is represented as a 32-bit number displayed as a
maximum of 8 hexadecimal characters.
ii. Port number: It is represented as 4 hexadecimal characters.
iii. A unique identifier for the instance of this application.
12.The authorization identifier.
13.The engine dispatchable unit identifier.
14.The name of the engine dispatchable unit.
15.The product name (“DB2”), component name (“data management”), and
function name (“sqlInitDBCB”) that is writing the message (as well as the
probe point (“4820”) within the function).
16.The information returned by a called function. There may be multiple data
fields returned.

Trap files
The database manger generates a trap file if it cannot continue processing due
to a trap, segmentation violation, or exception.

All signals or exceptions received by DB2 are recorded in the trap file. The trap
file also contains the function sequence that was running when the error
occurred. This sequence is sometimes referred to as the “function call stack” or

Chapter 10. Testing and tuning 337

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

“stack trace.” The trap file also contains additional information about the state of
the process when the signal or exception was caught.

A trap file is also generated when an application is forced while running a fenced
thread-safe routine. The trap occurs as the process is shutting down. This is not
a fatal error and it is nothing to be concerned about.

The files are located in the directory specified by the DIAGPATH database
manager configuration parameter.

On all platforms, the trap file name begins with a process identifier (PID),
followed by a thread identifier (TID), followed by the partition number (000 on
single partition databases), and concluded with .trap.txt.

There are also diagnostic traps, generated by the code when certain conditions
occur which do not warrant crashing of the instance, but where it may be useful
to verify values within the stack. These traps are named with the PID in decimal
format, followed by the partition number (0 in a single partition database).

The following resembles a trap file with a process identifier (PID) of 6881492,
and a thread identifier (TID) of 2.
6881492.2.000.trap.txt

The following is an example of a trap file whose process and thread is running on
partition 10.
6881492.2.010.trap.txt

You can generate trap files on demand using the db2pd command with the -stack
all or -dump option. In general, though, this should only be done as requested by
IBM Software Support.

Dump files
When DB2 determines that extra information is required for collection due to an
error, it often creates binary dump files in the diagnostic path. This file are named
with the process or thread ID that failed, the node where the problem occurred
and ends with .dump.bin extension, as shown in the example below.
6881492.2.010.dump.bin

When a dump file is created or appended, an entry is made in the db2diag log file
indicating the time and the type of data written.

Core files (Linux/UNIX)

If a program terminates abnormally, a core file is created by the system to store a
memory image of the terminated process. Errors such as memory address

338 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

violations, illegal instructions, bus errors, and user-generated quit signals cause
core files to be dumped.

These files are located in the directory specified by the DIAGPATH database
manager configuration parameter.

10.4.4 DB2 support information

Identifying what information is required to resolve problems is another important
step. All conditions define problem that are essential when reviewing
documentation, searching through available Internet resources, or contacting
DB2 support.

Maintenance version
The db2level utility can be used to check the current version of DB2. As
presented in Figure 10-2, the utility returns information about the installed
maintenance updates (FixPaks), the word length used by the instance (32-bit or
64-bit), the build date, and other code identifiers. It is recommended to
periodically check if the newest available FixPaks are installed. DB2
maintenance updates are freely available at:
ftp://ftp.software.ibm.com/ps/products/db2/fixes

Figure 10-2 Sample db2level output

db2support utility
The db2support utility is designed to automatically collect all DB2 and system
diagnostic data. This program generates information about a DB2 server,
including that related to configuration and system environment.

The output of the program is stored in a single compressed file named

db2support.zip, located in the directory specified as part of the db2support
command options.

Chapter 10. Testing and tuning 339

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

In one simple step, the tool can gather database manager snapshots,
configuration files, and operating system parameters, which should make the
problem determination quicker. Below is a sample call of the utility:
db2support . -d invent -c

The dot represents the current directory where the output file is stored. The rest
of the command options are not required and can be omitted. The -d and -c
clauses instruct the utility to connect to the invent database, and to gather
information about database objects such as table spaces, tables, or packages.

DB2 Technical Support site

An invaluable place to look when experiencing a problem is the DB2 Technical
Support site for Linux, Windows, and UNIX located on the Web at:
http://www.ibm.com/software/data/db2/support/db2_9/

The site has the most recent copies of documentation, knowledge base to search
for technical recommendations or DB2 defects, links for product updates, latest
support news, and other useful DB2 related links.

To find related problems, prepare words that describe the issues, such as the
commands that were run, symptoms, or tokens from the diagnostics messages.
You can use these as terms in the DB2 Knowledge Base. The Knowledge Base
offers an option to search through DB2 documentation, TechNotes, and DB2
defects (APARs). TechNotes is a set of recommendations and solutions for
specific problems.

Authorized Program Analysis Reports (APARs) are defects in DB2 code

discovered by customers that require a fix. APARs have unique identifiers and
are always specific to a particular version, but may affect multiple products in the
DB2 family running on multiple platforms. Fixes for APARs are provided through
DB2 FixPaks.

On the DB2 support site there is a possibility to search for closed, open, and
HIPER APARs. A closed status for APAR indicates that a resolution for a
problem has been achieved and included in a specific FixPak. Open APARs
represent DB2 defects that are currently being worked on or waiting to be
included in the next available FixPak. HIPER APARs (High-Impact or
PERvasive) are critical problems that should be reviewed to assess the potential
impact of staying at a particular FixPak level.

The DB2 Technical Support site offers e-mail notification of critical or pervasive
DB2 customer support issues including HIPER APARs and FixPak alerts. To
subscribe to it, follow the DB2 Alert link on the Technical Support main page.

You can also send DB2 for LUW questions to:askdata@ca.ibm.com

340 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Calling IBM support

If the problem appears too complex to solve on your own, you can contact the
IBM Software Support Center. In order to understand and resolve your support
service request in the most expedient way, it is important that to gather
information about the problem and have it on hand when talking to the software
specialist.

Guidelines and reference materials (which you may need when calling IBM
support), as well as the telephone numbers are available in the IBM Software
Support Guide at:
http://techsupport.services.ibm.com/guides/handbook.html

10.4.5 Monitoring tools

Tuning and troubleshooting a database can be a complex process. DB2 comes
with various tools, functions, and applications that make this task much easier.
One such tool is the DB2 monitoring utility, which can collect information on
many different system activities such as usage of buffer pools, locks held by
applications, sorts performed by system, activities on tables, connections,
transactions statistics or statements run on the system. There are two main
methods of monitoring:
򐂰 Monitor table functions
򐂰 Snapshot™ monitoring
򐂰 Event monitoring

Monitor table functions

Starting with DB2 Version 9.7, you can access monitor data through a
light-weight alternative to the traditional system monitor. Use monitor table
functions to collect and view data for systems, activities, or data objects.

Data for monitored elements is continually accumulated in memory and available

for querying. You can choose to receive data for a single object (for example,
service class A or table TABLE1) or for all objects.

When using these table functions in a database partitioned environment, you can
choose to receive data for a single partition or for all partitions. If you choose to
receive data for all partitions, the table functions returns one row for each
partition. Using SQL, you can sum the values across partitions to obtain the
value of a monitor element across partitions.

Monitor table functions can be broken up into three categories, depending on the
information they monitor:

Chapter 10. Testing and tuning 341

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

򐂰 System: The system monitoring perspective includes the complete volume of

work and effort expended by the data server to process application requests.
This perspective helps you determine what the data server is doing as a
whole as well as for particular subset of application requests. The following is
a list of table functions for retrieving system information:
– MON_GET_SERVICE_SUBCLASS
– MON_GET_SERVICE_SUBCLASS_DETAILS
– MON_GET_WORKLOAD
– MON_GET_WORKLOAD_DETAILS
– MON_GET_CONNECTION
– MON_GET_CONNECTION_DETAILS
– MON_GET_UNIT_OF_WORK
– MON_GET_UNIT_OF_WORK_DETAILS
򐂰 Activities: The activity monitoring perspective focuses on the subset of data
server processing related to executing activities. In the context of SQL
statements, the term activity refers to the execution of the section for a SQL
statement. The following table functions tare used to access current data for
activities:
– MON_GET_ACTIVITY_DETAILS
– MON_GET_PKG_CACHE_STMT
򐂰 Data objects: The data object monitoring perspective provides information
about operations performed on data objects such as tables, indexes, buffer
pools, table spaces, and containers. Use the following table functions to
access current details for data objects:
– MON_GET_BUFFERPOOL
– MON_GET_TABLESPACE
– MON_GET_CONTAINER
– MON_GET_TABLE
– MON_GET_INDEX

Example 10-12 shows an example of how one can use the MON_GET_TABLE
function to retrieve the rows read, inserted, updated, and deleted from all tables
in the ADMIN schema.

Example 10-12 Monitor table function

db2inst1@db2server:~> db2 "SELECT varchar(tabschema,20) as tabschema,
varchar(tabname,20) as tabname,
sum(rows_read) as total_rows_read,
sum(rows_inserted) as total_rows_inserted,
sum(rows_updated) as total_rows_updated,
sum(rows_deleted) as total_rows_deleted
FROM TABLE(MON_GET_TABLE('ADMIN','',-1)) AS t
GROUP BY tabschema, tabname

342 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

ORDER BY total_rows_read
DESC"

Snapshot monitoring
Snapshot monitoring describes the state of database activity at a particular point
in time when a snapshot is taken. Snapshot monitoring is useful in determining
the current state of the database and its applications. Taken at regular intervals,
they are useful for observing trends and foreseeing potential problems.

Snapshots can be taken from the command line, using custom API programs or
through SQL using table functions. Example 10-13 shows an extract from a
sample snapshot invoked from the command line.

Example 10-13 Example snapshot

db2inst1@db2server:~> db2 GET SNAPSHOT FOR DATABASE ON invent

Database Snapshot

Database name = INVENT

Database path =
/home/db2inst1/invent/db2inst1/NODE0000/SQL00001/
Input database alias = INVENT
Database status = Active
Catalog database partition number = 0

[...]

High water mark for connections = 9

Application connects = 767
Secondary connects total = 7
Applications connected currently = 1
Appls. executing in db manager currently = 0
Agents associated with applications = 7
Maximum agents associated with applications= 9
Maximum coordinating agents = 9

[...]

Buffer pool data logical reads = Not Collected

Buffer pool data physical reads = Not Collected
Buffer pool temporary data logical reads = Not Collected
Buffer pool temporary data physical reads = Not Collected

[...]

In the example above, the snapshot has collected database level information for
the INVENT database. Some of the returned parameters display point-in-time
values such as the number of currently connected applications:

Chapter 10. Testing and tuning 343

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Applications connected currently = 1

Some parameters represent cumulative values such as the number of connect

statements issued against the database:
Application connects = 767

Some parameters can contain historical values such as the maximum number of
concurrent connections that have been observed on the database:
High water mark for connections = 9

Cumulative or historical values are used to relate to the point in time during last
initialization of counters. The counters can be reset to zero by the RESET MONITOR
command, or by the appropriate DB2 event. With the mentioned Example 10-13
on page 343, database deactivation and activation reset all the database level
counters. Example 10-14 shows how to reset monitors for an entire instance and
for the specific database.

Example 10-14 Resetting snapshot monitor counters

db2 RESET MONITOR ALL
db2 RESET MONITOR FOR DATABASE invent

To optimize database performance in a default DB2 configuration, most of the

snapshot monitor elements are not collected. Due to this, in Example 10-12 on
page 342, the value Not Collected was displayed for the buffer pool statistics.
DB2 contains monitor switches to provide database administrators with the
option of constraining the collection of monitor elements. Current monitor
switches set for the session can be displayed from the command line with the
GET MONITOR SWITCHES command as shown in Example 10-15.

Example 10-15 Displaying monitor switches

db2inst1@db2server:~> db2 GET MONITOR SWITCHES

Monitor Recording Switches

Switch list for db partition number 0

Buffer Pool Activity Information (BUFFERPOOL) = OFF
Lock Information (LOCK) = OFF
Sorting Information (SORT) = OFF
SQL Statement Information (STATEMENT) = OFF
Table Activity Information (TABLE) = OFF
Take Timestamp Information (TIMESTAMP) = ON 09/11/2009 22:10:34.801069
Unit of Work Information (UOW) =
OFF

344 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

The monitor switches can be turned on at the instance level or the application
level. To switch the monitors at the instance level, modify the appropriate
database manager parameter. After modifying the DFT_MON_BUFPOOL
parameter, as shown in Example 10-16, all users with SYSMAINT, SYSCTRL, or
SYSADM authorities are able to collect buffer pool statistics on any database in
the instance.

Example 10-16 Updating monitor switches at instance level

db2 UPDATE DBM CFG USING DFT_MON_BUFPOOL ON

To switch the monitors at the application level, issue the UPDATE MONITOR
SWITCHES command using command line. The changes only are applicable to
that particular prompt window. Example 10-17 shows how to update the suitable
monitor switch for collecting buffer pool information.

Example 10-17 Updating monitor switches at application level

db2 UPDATE MONITOR SWITCHES USING BUFFERPOOL ON

The complete list of monitor switches and related database manager parameters
is presented on Table 10-2.

Table 10-2 List of monitor switches and related DBM parameters

Database manager Monitor switch Information provided
parameter

DFT_MON_BUFFERPOOL BUFFERPOOL Number of reads and

writes, time taken

DFT_MON_LOCK LOCK Number of locks held, number of

deadlocks

DFT_MON_SORT SORT Number of heaps used, overflows, sort

performance

DFT_MON_STMT STATEMENT Start/stop time,

statement identification

DFT_MON_TABLE TABLE Measure of activity (rows

read/written)

DFT_MON_UOW UOW (Unit Of Work) Start/end times,

completion status

DFT_MON_TIMESTAMP TIMESTAMP Timestamps for operations

Sample snapshots
The database manager snapshot (Example 10-18 on page 346) captures
information specific to the instance level. The information centers around the

Chapter 10. Testing and tuning 345

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

total amount of memory allocated to the instance and the number of agents that
are currently active on the system.

Example 10-18 Database manager snapshot

db2 GET SNAPSHOT FOR DATABASE MANAGER

The lock snapshot (Example 10-19) is useful in determining what locks an

application is currently holding and the locks other applications are waiting on.
The snapshot lists all applications on the system and the locks that each of these
is holding. Each lock and each application is given a unique identifier number.

Example 10-19 Lock snapshot

db2 GET SNAPSHOT FOR LOCKS ON invent

The table snapshot (Example 10-20) contains information on the usage and
creation of all tables. This information is useful in determining how much work is
being run against a table and how much the table data changes. This information
can be used to decide how your data should be laid out physically.

Example 10-20 Table snapshot

db2 GET SNAPSHOT FOR TABLES ON invent

The table space and buffer pool snapshots (Example 10-21) contain similar
information. The table space snapshot returns information regarding the layout of
the table space and the amount of space being used. The buffer pool snapshot
contains information on the amount of space currently allocated for buffer pools,
and space required when the database is next reset. Both snapshots contain a
summary of the way in which data is accessed from the database. This access
can be done from a buffer pool, directly from tables on disk, or through a direct
read or write for LOBs or LONG objects.

Example 10-21 Table space and buffer pool snapshots

db2 GET SNAPSHOT FOR TABLESPACES ON invent
db2 GET SNAPSHOT FOR BUFFERPOOLS ON invent

The dynamic SQL snapshot (Example 10-22 on page 347) is used extensively to
determine how well SQL statements are performing. This snapshot summarizes
the behavior of the different dynamic SQL statements that are run. The snapshot
does not capture static SQL statements, so anything that was pre-bound does
not show up in this list. The snapshot is an collection of the information
concerning the SQL statements. If a SQL statement is executed 102 times, then

346 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

there will be one entry, which will encapsulate a summary of the behavior of all
102 executions.

Example 10-22 Dynamic SQL snapshot

db2 GET SNAPSHOT FOR DYNAMIC SQL ON invent

Snapshot table functions

Authorized users can also capture snapshot monitoring information for an
instance by using snapshot table functions or snapshot administrative views. The
snapshot table functions allow requesting of data for specific database partitions,
globally aggregated data, or data from all database partitions. Some snapshot
table functions allow you to request data from all active databases. The snapshot
administrative views provide a simple means of accessing data for all database
partitions of the connected database.

Example 10-23 and Example 10-24 show how to get similar monitoring
information using the table functions and views as we did from Example 10-17
using the GET SNAPSHOT command. Example 10-23 demonstrates a query
that captures the snapshot of lock information for the currently connected
database. Example 10-24 is a query that captures a snapshot of lock information
about the SAMPLE database for the current connected database partition.

Example 10-23 Sample snapshot table function

db2inst1@db2server:~> SELECT * FROM SYSIBMADM.SNAPLOCK

Example 10-24 Sample snapshot table function

db2inst1@db2server:~>SELECT * FROM TABLE(SNAP_GET_LOCK('invent',-1)) AS
SNAPLOCK

Table 10-3 lists the snapshot table functions, administrative views and return
information that can be used to monitor your database system. All administrative
views belong to the SYSIBMADM schema.

Table 10-3 Common snapshot table functions and administrative views

Monitor Snapshot table function Administrative View Information returned
level

Chapter 10. Testing and tuning 347

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

DBM SNAP_GET_DBM_V95 SNAPDBM Retrieve the database manager level

information.
SNAP_GET_FCM SNAPFCM Retrieve the fcm logical data group
snapshot information
SNAP_GET_FCM_PART SNAPFCM_PART Retrieve the fcm_node logical data group
snapshot information
SNAP_GET_SWITCHES SNAPSWITCHES Retrieve database snapshot switch state
information
SNAP_GET_DBM_MEMOR SNAPDBM_MEMORY_ Retrieve database manager level memory
Y_POOL POOL usage information
DB SNAP_GET_DB_V95 SNAPDB Retrieve snapshot information from the
database logical group
SNAP_GET_DB_MEMORY_ SNAPDB_MEMORY_ Retrieve database level memory usage
POOL POOL information
SNAP_GET_HADR SNAPHADR Retrieve HADR logical data group
snapshot information
APP SNAP_GET_APPL_V95 SNAPAPPL Retrieve application logical data group
snapshot information
SNAP_GET_APPL_INFO_ SNAPAPPL_INFO Retrieve appl_info logical data group
V95 snapshot information
SNAP_GET_LOCKWAIT SNAPLOCKWAIT Retrieve lockwait logical data group
snapshot information
SNAP_GET_STMT SNAPSTMT Retrieve statement snapshot information
SNAP_GET_AGENT SNAPAGENT Retrieve agent logical data group
application snapshot information
SNAP_GET_SUBSECTION SNAPSUBSECTION Retrieve subsection logical monitor group
snapshot information
SNAP_GET_AGENT_ SNAPAGENT_MEMORY Retrieve memory_pool logical data group
MEMORY_POOL _POOL snapshot information
Table SNAP_GET_TAB_V91 SNAPTAB Retrieve table logical data group snapshot
information
SNAP_GET_TAB_REORG SNAPTAB_REORG Retrieve table reorganization snapshot
information
Lock SNAP_GET_LOCK SNAPLOCK Retrieve lock logical data group snapshot
information

348 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Table SNAP_GET_TBSP_V91 SNAPTBSP Retrieve table space logical data group

space snapshot information
SNAP_GET_TBSP_PART_ SNAPTBSP_PART Retrieve tablespace_nodeinfo logical data
V91 group snapshot information
SNAP_GET_TBSP_ SNAPTBSP_QUIESCER Retrieve quiescer table space snapshot
QUIESCER information
SNAP_GET_CONTAINER_ SNAPCONTAINER Retrieve tablespace_container logical data
V91 group snapshot information
SNAP_GET_TBSP_RANGE SNAPTBSP_RANGE Retrieve range snapshot information
Buffer SNAP_GET_BP_V95 SNAPBP Retrieve buffer pool logical group snapshot
pool information
SNAP_GET_BP_PART SNAPBP_PART Retrieve bufferpool_nodeinfo logical data
group snapshot information
Dynamic SNAP_GET_DYN_SQL_V95 SNAPDYN_SQL Retrieve dynsql logical group snapshot
SQL information
DB SNAP_GET_UTIL SNAPUTIL Retrieve utility_info logical data group
snapshot information
SNAP_GET_UTIL_PROGRE SNAPUTIL_PROGRESS Retrieve progress logical data group
SS snapshot information
SNAP_GET_DETAILLOG_ SNAPDETAILLOG Retrieve snapshot information from the
V91 detail_log logical data group
SNAP_GET_STORAGE_ SNAPSTORAGE_PATHS Retrieve automatic storage path
PATHS information

The SQL table functions have two input parameters:

򐂰 database name
VARCHAR(255). If you enter NULL, the name of the currently connected 1
database is used. 1
򐂰 partition number
SMALLINT. For the partition number parameter, enter the integer (a value 1
between 0 and 999) corresponding to the partition number you must monitor.
1 To capture a snapshot for the currently connected partition, enter a value 1
of -1 or a NULL. To capture a global snapshot, enter a value of -2.

For the following list of snapshot table functions, if you enter NULL for the
currently connected database, you get snapshot information for all databases in
the instance:
򐂰 SNAP_GET_DB_V95
򐂰 SNAP_GET_DB_MEMORY_POOL
򐂰 SNAP_GET_DETAILLOG_V91
򐂰 SNAP_GET_HADR

Chapter 10. Testing and tuning 349

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

򐂰 SNAP_GET_STORAGE_PATHS
򐂰 SNAP_GET_APPL_V95
򐂰 SNAP_GET_APPL_INFO_V95
򐂰 SNAP_GET_AGENT
򐂰 SNAP_GET_AGENT_MEMORY_POOL
򐂰 SNAP_GET_STMT
򐂰 SNAP_GET_SUBSECTION
򐂰 SNAP_GET_BP_V95
򐂰 SNAP_GET_BP_PART

The database name parameter does not apply to the database manager level
snapshot table functions; they have an optional parameter for database partition
number.

Event monitoring
Event monitors are used to monitor the performance of DB2 over a fixed period
of time. The information that can be captured by an event monitor similar to the
snapshots, but in addition to snapshot-level information, event monitors also
examine transition events in the database, and consider each event as an object.
Event monitors can capture information about DB2 events in the following areas:
򐂰 Statements: A statement event is recorded when an SQL statement ends. The
monitor records statement's start and stop time, CPU used, text of dynamic
SQL, the return code of the SQL statement, and other matrixes such as fetch
count.
򐂰 Connections: A connection event is recorded whenever an application
disconnects from the database. The connection event records all application
level counters.
򐂰 Database: An event of database information is recorded when the last
application disconnects from the database. This event records all database
level counters.
򐂰 Buffer pools; A buffer pool event is recorded when the last application
disconnects from the database. The information captured contains the type
and volume of use of the buffer pool, use of pre-fetchers and page cleaners,
and whether or not direct I/O was used.
򐂰 Table spaces: A table space event is recorded when the last application
disconnects from the database. This monitor captures information regarding
counters for buffer pool, prefetchers, page cleaners and direct I/O for each
table space.
򐂰 Tables: All active table events are recorded when the last application
disconnects from the database. An active table is one which has been altered
or created since the database was activated. The monitor captures the
number of rows read and written to the table.

350 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

򐂰 Activities: An activity event is recorded on the completion of an activity that

executed in a service class, workload or work class that has had its
COLLECT ACTIVITY DATA option turned on. Data is also collected for the
targeted activity in the instant the
WLM_CAPTURE_ACTIVITY_IN_PROGRESS stored procedure is executed.
Data can also be collected if the activity violates a threshold that has the
COLLECT ACTIVITY DATA option enabled. The activity event monitor
records activity-level data. If WITH DETAILS option was specified as part of
the COLLECT ACTIVITY DATA statement, this includes statement and
compilation environment information for those activities that have it. If AND
VALUES was also specified, this also includes input data values for those
activities that have it.
򐂰 Statistics: A statistics event is recorded every xx minutes, where xx is the
length of time over which statistics are gathered. This period is defined in the
WLM_COLLECT_INT database configuration parameter. Data can also be
collected when the WLM_COLLECT_STATS stored procedure is called. The
statistics captured are activities that executed with each service class,
workload, or work class that exists on the system.
򐂰 Threshold violations: A Threshold violation event is recorded upon detection
of a threshold violation. The threshold violations event monitor records a
description of the threshold that was violated (the identifier, maximum value,
and so on), identification of the activity that violated the threshold and the unit
of work identifier and he time that the threshold was violated.
򐂰 Locking: The locking event monitor is new and can replace the deprecated
deadlock event monitor. A locking event is recorded upon detection of any of
the following event types, depending on the configuration: lock timeout,
deadlock, lock wait beyond a specified duration. The information captured
from this event monitor focuses on the locks involved in the failure and the
applications that own them.
򐂰 Unit of work: The unit of work is a new event monitoring type to DB2 9.7 and
it is to replace the deprecated transaction event monitor. A unit of work event
is recorded upon completion of a unit of work. The Unit of work event records
a variety of information, including attributes at the database level, connection
level, and unit of work level.

Event monitors are created with the CREATE EVENT MONITOR SQL
statement. Information about event monitors is stored in the system catalog
table, and it can be reused later.

Example 10-25 on page 352 shows a sequence of statements that illustrate how
to collect Event Monitor information using commands.

Chapter 10. Testing and tuning 351

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Example 10-25 Working with event monitors

1 db2inst1@db2server:~> db2 "CREATE EVENT MONITOR mymon1 FOR DATABASE,
STATEMENTS WRITE TO FILE '/home/db2inst1/temp'"
DB20000I The SQL command completed successfully.

2 db2inst1@db2server:~> db2 SET EVENT MONITOR mymon1 STATE=1

3 db2inst1@db2server:~/> db2 "SELECT * FROM amdin.services"

[…]

4 db2inst1@db2server:~/> db2 SET EVENT MONITOR mymon1 STATE=0

5 db2inst1@db2server:~/> db2 DROP EVENT MONITOR mymon1

DB20000I The SQL command completed successfully.

6 db2inst1@db2server:~/> db2evmon -path /home/db2inst1/temp

The numbers bolded in the example correspond to the following descriptions:

1. This line is the statement used to create the event monitor mymon1. which
specifically collects DATABASE and STATEMENT events.
2. This line is turning the event monitor on. Values for this statement can be 0
for off and 1 for on.
3. This line is used as a SQL statement that should be captured by the event
monitor.
4. This line turns off the event monitor.
5. This line drops the event monitor mymon1.
6. The db2evmon utility converts the event monitor binary files to a user readable
form.Example 10-26 shows the output from this utility.

Example 10-26 db2evemon output

-------------------------------------------------------------------------
EVENT LOG HEADER
Event Monitor name: MYMON1
Server Product ID: SQL09070
Version of event monitor data: 10
Byte order: LITTLE ENDIAN
Number of nodes in db2 instance: 1
Codepage of database: 1208
Territory code of database: 1
Server instance name: db2inst1
------------------------------------------------------------------------
------------------------------------------------------------------------
Database Name: INVENT
Database Path: /home/db2inst1/invent/db2inst1/NODE0000/SQL00001/

352 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

First connection timestamp: 09/16/2009 15:16:33.138148

Event Monitor Start time: 09/17/2009 10:12:14.573251
------------------------------------------------------------------------

3) Connection Header Event ...

Appl Handle: 4195
Appl Id: *LOCAL.db2inst1.090917034642
Appl Seq number: 00029
DRDA AS Correlation Token: *LOCAL.db2inst1.090917034642
Program Name : db2bp
Authorization Id: DB2INST1
Execution Id : db2inst1
Codepage Id: 1208
Territory code: 1
Client Process Id: 12152
Client Database Alias: INVENT
Client Product Id: SQL09070
Client Platform: Unknown
Client Communication Protocol: Local
Client Network Name: db2server
Connect timestamp: 09/16/2009 23:46:42.964073

[...]

10.4.6 Visual Explain

An access plan is a cost estimation of resource usage for a query, based on
available information such as statistics for tables and indexes, instance and
database configuration parameters, bind options and query optimization level,
and so on. An access plan also specifies the order of operations for accessing
the data.

The access plan acquired from Visual Explain helps to understand how individual
SQL or XQuery statements are executed. The information displayed in the Visual
Explain graph can be used to tune SQL and XQuery queries to optimize
performance.

You can start Visual Explain from the Control Center or the Optim Data Studio
toolset. From Data Studio create or open SQL or XQuery statement. In the Main
panel view right click and select Open Visual Explain, as shown in Figure 10-3
on page 354.

Chapter 10. Testing and tuning 353

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Figure 10-3 Opening Visual Explain in Data Studio

A configuration window appears where you can specify the general settings and
the values for Visual Explain to use for special registers when fetching explain
data.

Figure 10-4 on page 355 shows an example of an access plan graph. To get the
details right-click the desired graph element.

354 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Figure 10-4 Visual Explain access plan graph

You can use Visual Explain to:

򐂰 View the statistics that were used at the time of optimization. You can
compare these statistics to the current catalog statistics to help you determine
whether rebinding the package might improve performance.
򐂰 Determine whether or not an index was used to access a table. If an index
was not used, Visual Explain can help you determine which columns might
benefit from being indexed.
򐂰 View the effects of performing various tuning techniques by comparing the
before and after versions of the access plan graph for a query.
򐂰 Obtain information about each operation in the access plan, including the
total estimated cost and number of rows retrieved (cardinality).

10.5 Initial tuning

Performance of a DB2 database application can be influenced by many factors
such as the type of workload, application design, database design, capacity
planning, and instance and database configuration. Initial tuning has become
easier with the introduction of autonomic computing and most automatic features

Chapter 10. Testing and tuning 355

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

which are turned on by default with DB2. However, there are still a few settings
you may required to modify to fit your environment. This section focuses on a
number of DB2 performance tuning tips that may be used for initial configuration.

10.5.1 Table spaces design

At database creation time, three table spaces are created:
򐂰 SYSCATSPACE - Catalog table space for storing information about all the
objects in the database.
򐂰 TEMPSPACE1 - System temporary table space for storing internal temporary
data required during SQL operations such as sorting, reorganizing tables,
creating indexes, and joining tables.
򐂰 USERSPACE1 - For storing user-defined tables.

By default SYSCATSPACE and USERSPACE1 table spaces are created as

Database Managed Space (DMS), which means that the database manager
controls the storage space. The TEMPSPACE1 table space is created as System
Managed Space (SMS), meaning that the regular operating system functions are
used for handling I/O operations.

For very simple databases, the default configurations may be sufficient for your
needs. However, in most cases you may want to add additional table spaces.
The benefit of multiple table spaces is that you can assign different database
objects to different table spaces and assign the table spaces to dedicated
physical devices. Therefore allowing each table object to utilize the hardware
allocated to the table space it belongs to. This essentially allows for table level
backup.

You can create an SMS tables space using the MANAGED BY SYSTEM clause
in the create table space definition. The benefit of using a SMS table space is
that it does not require initial storage; space is not allocated by the system until it
is required. Creating a table space with SMS requires less initial work, because
you do not have to predefine the containers.

However, with SMS table spaces, the file system of the operating system
decides where each logical file page is physically stored. This means that pages
may not be stored contiguously on disk - the storage placement depends on the
file system algorithm and on the level of activity on the file system. This may
cause the performance of an SMS table space to be negatively affected.
Therefore, SMS table spaces are ideal for small database that require low
maintenance/monitoring and grow/shrink rapidly.

With DMS, the database manager can ensures that pages are physically
contiguous, since it bypasses operating system I/O and interfaces with the disk

356 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

directly. This can improve performance significantly. You can create a DMS
tables space by using MANAGED BY DATABASE clause in the create table
space definition.

The down side is that a DMS table space requires more efforts in tuning and
administration, since you must add more storage containers as the table space
fills with data. However, you can easily add new containers, drop, or modify the
size of existing containers. The database manager then automatically
rebalances existing data into all the containers belonging to the table space.
Therefore, DMS table spaces are idea for performance-sensitive application,
particularly ones that involve a large number of INSERT operations.

If a database is enabled for automatic storage (which is enabled by default),

there is a third option when creating a table space. Automatic management may
be specified using MANAGED BY AUTOMATIC STORAGE clause in the
CREATE TABLESPACE definition. With this option DB2 decides what type of
table space to create. There is no need to specify container details because DB2
assigns the containers and manages table space creation automatically.

For optimal performance, large volume data and indexes should be placed within
DMS table spaces; if possible, split to separate raw devices. Initially, system
temporary table spaces should be of SMS type. In an OLTP environment, there
is no need for creating large temporary objects to process SQL queries, so the
SMS system temporary table space is a good starting point. The easiest way to
optimize your table spaces is to use table spaces managed by automatic
storage.

10.5.2 Physical placement of database objects

When creating a database, the first important decision is the storage
architecture. The ideal situation is to have the fastest disks possible and at least
five to ten disks per processor (for high I/O OLTP workload use even more). The
reality is hardware is often chosen based on other considerations, so in order to
achieve optimal performance, the placement of database objects should be
carefully planned.

Chapter 10. Testing and tuning 357

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Figure 10-5 Explaining logical log

As shown in Figure 10-5, all data modifications are not only written to table space
containers, but are also logged to ensure recoverability. Since every INSERT,
UPDATE, or DELETE statement is replicated in the transactional log, the
flushing speed of the logical log buffer can be crucial for the entire database
performance. To understand the importance of logical log placement, you should
keep in mind that the time necessary to write data to disk depends on the
physical data distribution on disk. The more random reads or writes that are
performed, the more disk head movements are required, and therefore, the
slower the writing speed. Flushing logical log buffer to disk by its nature is
sequential and should not be interfered by other operations. Locating logical log
files on separate devices isolates them from other processes, and ensures
uninterrupted sequential writes.

To change logical log files to a new location you must modify the NEWLOGPATH
database parameter as shown in Example 10-27. The logs are relocated to the
new path on the next database activation (this can take some time to create the
files).

Example 10-27 Relocation of logical logs

db2 UPDATE DB CFG FOR SAMPLE USING NEWLOGPATH /db2/logs

When creating a DMS table space with many containers, DB2 automatically
distributes the data across them in a round-robin fashion, similar to the striping

358 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

method available in disk arrays. To achieve the best possible performance, each
table space container should be placed on a dedicated physical device. For
parallel asynchronous writes and reads from multiple devices, the number of
database page cleaners (NUM_IO_CLEANERS) and I/O servers
(NUM_IOSERVERS) should be adjusted. The best value for these two
parameters depends on the type of workload and available resources. You can
start your configuration with the following values:
򐂰 NUM_IOSERVERS = Number of physical devices, not less than three and no
more than five times the number of CPUs.
򐂰 NUM_IO_CLEANERS = Number of CPUs

However, the most effective way to configure these parameters is to set them to
automatic and let DB2 manage them, as shown in Example 10-28.

Example 10-28 Updating IO related processes

db2 UPDATE DB CFG FOR sample USING NUM_IOSERVERS AUTOMATIC
db2 UPDATE DB CFG FOR sample USING NUM_IOCLEANERS AUTOMATIC

If there is a relatively small number of disks available, it can be difficult to keep

logical logs, data, indexes, system temporary table spaces (more important for
processing large queries in a warehousing environment), backup files, or the
operating system paging file on separate physical devices. A compromise
solution is to have one large file system striped by a disk array (RAID device) and
create table spaces with only one container. The load balancing is shifted to
hardware, and you do not have to worry about space utilization. If you want
parallel I/O operations on a single container, the DB2_PARALLEL_IO registry
variable should be set before starting the DB2 engine.

If this registry variable is set, and the prefetch size of the table is not
AUTOMATIC, the degree of parallelism of the table space is the prefetch size
divided by the extent size. If this registry variable is set, and the prefetch size of
the table space is AUTOMATIC, DB2 automatically calculates the prefetch size
of a table space. Table 10-4 summarizes the different options available and how
parallelism is calculated for each situation:

Table 10-4 Calculating parallelism

Prefetch size of table space DB2_PARALLEL_IO Setting Parallelism is equal to:

AUTOMATIC Not set Number of containers

AUTOMATIC Table space ID Number of containers * 6

AUTOMATIC Table space ID:n Number of containers * n

Not AUTOMATIC Not set Number of containers

Not AUTOMATIC Table space ID Prefetch size/extent size

Chapter 10. Testing and tuning 359

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Not AUTOMATIC Table space ID:n Prefetch size/extent size

10.5.3 Buffer pools

The default size for buffer pools is relatively small: only 250 pages (~ 1 MB) for
Windows and 1000 pages (~ 4 MB) for Linux and UNIX platforms. The overall
buffer size has a great effect on DB2 performance by significantly reducing I/O,
the most time consuming operation. It is recommended to increase these default
values. However, the total buffer pool size should not be set too high, because
there might not be enough memory to allocate the desired size. To calculate the
maximum buffer size, all other DB2 memory related parameters such as
database heap, the agent's memory, storage for locks, as well as the operating
system and any other applications should be considered.

Initially, set the total size of buffer pools to 10% to 20% of available memory. You
can monitor the system later and correct it. DB2 allows changing buffer pool
sizes without shutting down the database. The ALTER BUFFERPOOL statement
with the IMMEDIATE option takes effect right away, except when there is not
enough reserved space in the database-shared memory to allocate new space.
This feature can be used to tune database performance according to periodical
changes in use, for example, switching from daytime interactive use to nighttime
batch work.

Once the total available size is determined, this area can be divided into different
buffer pools to improve utilization. Having more than one buffer pool can
preserve data in the buffers. For example, let us suppose that a database has
many frequently used smaller tables, which normally reside in the buffer in their
entirety, and thus would be accessible quickly. Now let us suppose that there is a
query running against a very large table using the same buffer pool and involving
reads of more pages than total buffer size. When this query runs, the pages from
the small, very frequently used tables will be lost, making it necessary to re-read
them when they are needed again.

At the start you can create additional buffer pools for caching data and leave the
IBMDEFAULTBP for system catalogs. Creating an extra buffer pool for system
temporary data can be also valuable for the system performance, especially in
an OLTP environment where the temporary objects are relatively small. Isolated
temporary buffer pools are not influenced by the current workload, so it should
take less time to find free pages for temporary structures, and it is likely that the
modified pages will not be swapped out to disk. In a warehousing environment,
the operation on temporary table spaces are considerably more intensive, so the
buffer pools should be larger, or combined with other buffer pools if there is not
enough memory in the system (one pool for caching data and temporary
operations).

360 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Example 10-29 shows how to create buffer pools assuming there is an additional
table space DATASPACE for storing data and indexes was already created and
that there is enough memory in the system. You can take this as a starting buffer
pool configuration for a 2 GB RAM system.

Example 10-29 Increasing buffer pools

connect to sample;
-- creating two buffer pools 256 MB and 64 MB
CREATE BUFFERPOOL data_bp IMMEDIATE SIZE 65536 pagesize 4k;
CREATE BUFFERPOOL temp_bp IMMEDIATE SIZE 16384 pagesize 4k;

-- changing size of the default buffer pool

ALTER BUFFERPOOL ibmdefaultbp IMMEDIATE SIZE 16384;

-- binding the tablespaces to buffer pools

ALTER TABLESPACE dataspace BUFFERPOOL data_bp;
ALTER TABLESPACE tempspace1 BUFFERPOOL temp_bp;

-- checking the results

SELECT
substr(bs.bpname,1,20) AS bpname
,bs.npages
,bs.pagesize
,substr(ts.tbspace,1,20) as TBSPACE
FROM syscat.bufferpools bs JOIN syscat.tablespaces ts ON
bs.bufferpoolid = ts. bufferpoolid;

The results:
BPNAME NPAGES PAGESIZE TBSPACE
-------------------- ----------- ----------- --------------------
IBMDEFAULTBP 16384 4096 SYSCATSPACE
IBMDEFAULTBP 16384 4096 SYSTOOLSPACE
IBMDEFAULTBP 16384 4096 USERSPACE1
DATA_BP 65536 4096 DATASPACE
TEMP_BP 16384 4096 TEMPSPACE1

Though you can tune your buffer pools manually as described above, using
STMM is an easier and more effective way of tuning the buffer pools for optimal
performance. As we discussed in 9.6, “Autonomics” on page 306, STMM can
tune database memory parameters and buffer pools without any DBA
intervention. STTM works with buffer pools of multiple page sizes and can easily
trade memory between the buffer pools as needed. You can turn on STMM for a
specific buffer pool by issuing commands in Example 10-30 on page 362.

Chapter 10. Testing and tuning 361

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Example 10-30 Selft-tuning memory manager and tuning buffer pools

db2inst1@db2server:~ > db2 UPDATE DB CFG FOR DATABASE invent USING
SELF_TUNING_MEM ON
db2inst1@db2server:~ > db2 ALTER BUFFERPOOL bp32 SIZE AUTOMATIC

The first command in the example above turns STMM on, which is the default.
The second command tells DB2 to automatically tune the buffer pool BP32. You
can tune individual buffer pools or all buffer pools with STMM.

The CHNGPGS_THRESH parameter specifies the percentage of changed

pages at which the asynchronous page cleaners will be started. Asynchronous
page cleaners write changed pages from the buffer pool to disk. The default
value for the parameter is 60%. When that threshold is reached, some users may
experience a slower response time. Having larger buffer pools means more
modified pages in memory and more work to be done by page cleaners, as
shown on Figure 10-6.

Figure 10-6 Visualizing CHNGPGS_THRESH parameter

For databases with a heavy update transaction workload, you can generally
ensure that there are enough clean pages in the buffer pool by setting the
parameter value to be equal-to or less-than the default value. A percentage
larger than the default can help performance of your database if there are a small
number of very large tables. To change the default parameter, you can use the
following command:
db2 update db cfg for sample using CHNGPGS_THRESH 40

362 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

10.5.4 Large transactions

By default, databases are created with relatively small space for transactional
logs, only three log files with 250 pages each, on Windows and 1000 pages on
Linux and UNIX.

A single transaction should fit into the available log space to be completed; if it
does not fit, the transaction is rolled back by the system (SQL0964C The
transaction log for the database is full). To process transactions, which are
modifying large numbers of rows, adequate log space is needed.

The currently total log space available for transactions can be calculated by
multiplying the size of one log file (database parameter LOGFILSIZ) and the
number of logs (database parameter LOGPRIMARY).

From a performance perspective, it is better to have a larger log file size because
of the cost for switching from one log to another. When log archiving is switched
on, the log size also indicates the amount of data for archiving. In this case, a
larger log file size is not necessarily better, since a larger log file size may
increase the chance of failure, or cause a delay in archiving or log shipping
scenarios. The log size and the number of logs should be balanced.

The following Example 10-31 allocates 400 MB of total log space.

Example 10-31 Resizing the transactional log

db2 UPDATE DB CFG FOR sample USING logfilsiz 5120
db2 UPDATE DB CFG FOR sample USING logprimary 20

Locking is the mechanism that the database manager uses to control concurrent
access to data in the database by multiple applications. Each database has its
own list of locks (a structure stored in memory that contains the locks held by all
applications concurrently connected to the database). The size of the lock list is
controlled by the LOCKLIST database parameter.

The default storage for LOCKLIST on Windows and UNIX is set to AUTOMATIC.
On 32-bit platforms, each lock requires 48 or 96 bytes of the lock list, depending
on whether other locks are held on the object or not. On 64-bit platforms, each
lock requires 64 or 128 bytes of the lock list, depending on whether other locks
are held on the object or not.

When this parameter is set to AUTOMATIC, it is enabled for self tuning. This
allows the memory tuner to dynamically size the memory area controlled by this
parameter as the workload requirements change. Since the memory tuner trades
memory resources among different memory consumers, there must be at least

Chapter 10. Testing and tuning 363

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

two memory consumers enabled for self tuning in order for self tuning to be
active.

The value of LOCKLIST is tuned together with the MAXLOCKS parameter.

Therefore, disabling self tuning of the LOCKLIST parameter automatically
disables self tuning of the MAXLOCKS parameter. Enabling self tuning of the
LOCKLIST parameter automatically enables self tuning of the MAXLOCKS
parameter.

Automatic tuning of this configuration parameter only occurs when self tuning
memory is enabled for the database (the SELF_TUNING_MEM database
configuration parameter is set to ON).

When the maximum number of lock requests has been reached, the database
manager replaces existing row-level locks with table locks (lock escalation). This
operation reduces the requirements for lock space, because transactions will
hold only one lock on the entire table instead of many locks on every row. Lock
escalation has a negative performance impact because it reduces concurrency
on shared objects. Other transactions must wait until the transaction holding the
table lock commits or rollbacks work. Setting LOCKLIST to AUTOMATIC avoids
this situation, as the lock list will increase synchronously to avoid lock escalation
or a “lock list ful” situation.

To check current usage of locks, use snapshots such as in Example 10-32.

Example 10-32 Invoking snapshot for locks on database invent

db2 get snapshot for locks on invent

The snapshot collects the requested information at the time the command was
issued. Issuing the get snapshot command later can produce different results
because in the meantime the applications may commit the transaction and
release locks.To check lock escalation occurrences, look at the db2diag.log file.

Log buffer
Log records are written to disk when one of the following occurs:
򐂰 A transaction commits or a group of transactions commit, as defined by the
mincommit configuration parameter
򐂰 The log buffer is full
򐂰 As a result of some other internal database manager event.

This log buffer size must also be less than or equal to the dbheap parameter.
Buffering the log records results in more efficient logging file I/O because the log

364 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

records are written to disk less frequently and a greater quantity of log records
are written out at each time.

The default size for log buffer is 256 4KB pages. In most cases the log records
are written to disk when one of the transactions issue a COMMIT, or the log
buffer is full. It is recommended to increase the size of this buffer area if there is
considerable read activity on a dedicated log disk, or there is high disk utilization.
Increasing the size of the log buffer may result in more efficient I/O operations,
especially when the buffer is flushed to disk. The log records are written to disk
less frequently and more log records are written each time.

When increasing the value of this parameter, you should also consider
increasing the DBHEAP parameter since the log buffer area uses space
controlled by the DBHEAP parameter.

At a later time, you can use the snapshot for applications to check the current
usage of log space by transactions as presented in Example 10-33.

Example 10-33 Current usage of log space by applications

$db2 UPDATE MONITOR SWITCHES USING UOW ON
$db2 GET SNAPSHOT FOR APPLICATIONS ON sample | grep "UOW log"

UOW log space used (Bytes) = 478

UOW log space used (Bytes) = 21324
UOW log space used (Bytes) = 110865

Before running the application snapshot, the Unit Of Work monitor should be
switched on. At the time the snapshot was issued, you can see that there are
only three applications running on the system. The first transaction uses 478
bytes of log space, the second 21324, and last 110865, which is roughly 28
pages more than the default log buffer size. The snapshot gives only current
values from the moment the command was issued. To get more valuable
information about the usage of log space by transactions, run the snapshot many
times.

10.5.5 SQL execution plan

DB2 prepares an access plan, whenever a SQL or XQuery statement is issued
against the database. This plan specifies the order of operation for accessing the
data. The access plan is created at prep/bind-time for static statements, and at
run-time for dynamic statements.

An access plan can be used to view statistics for selected tables, indexes, or
columns; properties for operators; global information such as table space and
function statistics; and configuration parameters relevant to optimization. With

Chapter 10. Testing and tuning 365

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Visual Explain, you can view the access plan for an SQL or XQuery statement in
graphical form.

It is important to understand that an access plan is an estimate based on the

information that is available. The optimizer bases its estimations on information
such as the following:
򐂰 Statistics in system catalog tables (if statistics are not current, update them
using the RUNSTATS command.)
򐂰 Configuration parameters
򐂰 Bind options
򐂰 The query optimization class

After many changes to table data, logically sequential data might reside on
non-sequential data pages, so that the database manager must perform
additional read operations to access data.

Additional read operations are also required if many rows have been deleted. In
this case, you may consider reorganizing the table to match the index and to
reclaim space. You can also reorganize the system catalog tables.

Because reorganizing a table usually takes more time than updating statistics,
you could execute the RUNSTATS command to refresh the current statistics for
your data, and then rebind your applications. If refreshed statistics do not
improve performance, reorganization might help.

The RUNSTATS command can be executed against a table from the command
line. Example 10-34 shows how to execute RUNSTATS command against our
sample inventory table.

Example 10-34 Executing RUNSTATS on the inventory table

db2inst1@db2server:~> db2 "RUNSTATS ON TABLE ADMIN.INVENTORY"
DB20000I The RUNSTATS command completed successfully.

It is also possible to update statistics using the Data Studio tool. Within the
Database Explorer View, connect to your database and drill down the database
object folders until you find the table you would like to update the statistics for. In
our example we connected to the invent database, then drop down the invent
database folder, the schema folder, the ADMIN schema folder, the Tables folder
and this brought us to a list of tables in the invent database in the ADMIN
schema. To pull up the table options we right clicked the INVENTORY table
icon, as shown in illustrated in Figure 10-7 on page 367.

366 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Figure 10-7 Running RUNSTATS on the INVENT table

The output of the update statistics is displayed in Figure 10-8.

Figure 10-8 Output of RUNSTATS on the INVENT table

Chapter 10. Testing and tuning 367

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

In some scenarios you may need more than current statistics on a table to
improve performance. The following factors can indicate if your database will
benefit from table reorganization:
򐂰 There has been a high volume of insert, update, and delete activity against
tables that are accessed by queries.
򐂰 There have been significant changes in the performance of queries that use
an index with a high cluster ratio.
򐂰 Executing the RUNSTATS command to refresh table statistics does not
improve performance.
򐂰 Output from the REORGCHK command indicates a need for table
reorganization.

You can access the table reorganization option from the Data Studio by
right-clicking on the table, as shown in Figure 10-7 on page 367. After selecting
the REORG Table... option in the drop down menu, the reorganization table
wizard opens in the main view. Figure 10-9 on page 369 illustrates the
reorganization table wizard. You can select the parameters for the REORG
command and select Run to execute the command.

368 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Figure 10-9 Reorganization table wizard

It can be time consuming to constantly monitor the database to see when

statistics must be refreshed. As discussed in 9.6, “Autonomics” on page 306,
DB2 autonomics can make things easier for you. With autonomics, you can
create a schedule for DB2, to perform regular table or index reorganization,
statistics collection and profiling. Figure 10-10 on page 370 shows configure
automatic maintenance wizard within the Data Studio GUI tool. To open this
wizard, right-click on the database you would like to configure automatic
maintenance for and select the Update statistics option, as higlighted in figure
Figure 10-7 on page 367.

Chapter 10. Testing and tuning 369

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Figure 10-10 Automatic maintenance wizard

DB2 comes with a very powerful query optimization algorithm. This cost-based
algorithm will attempt to determine the cheapest way to perform a query against
a database. Items such as the database configuration, database physical layout,
table relationships, and data distribution are all considered when finding the
optimal access plan for a query. To check the current execution plan, you can
use the Explain utility.

10.5.6 Configuration Advisor

The Configuration Advisor wizard is a tool that can be helpful in preparing the
DB2 initial configurations. The wizard requests information about the database,
its data, and the purpose of the system, and then recommends new configuration
parameters for the database and the instance. It can be run manually, through a
GUI or command line interface, and automatically during database creation.

By default in DB2 9, any new databases will run the Configuration Advisor in the
background and have such configuration recommendations automatically
applied. To disable this feature, or to explicitly enable it, you must use the db2set
command, as Example 10-35 illustrates.

370 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Example 10-35 Configuring default behavior of the Configuration Advisor

db2set DB2_ENABLE_AUTOCONFIG_DEFAULT=NO
db2set DB2_ENABLE_AUTOCONFIG_DEFAULT=YES

In any case, the Configuration Advisor can be manually run at any time against a
database to update the current configuration, regardless of the
DB2_ENABLE_AUTOCONFIG_DEFAULT setting. All recommendations are
based on the input that you provide and system information that the
Configuration Advisor gathers. The generated recommendations can be applied
or simply displayed.

It is important to point out that the values suggested by the Configuration Advisor
are relevant for only one database per instance. If you want to use this advisor
on more than one database, each database should belong to a separate
instance.

The Configuration Advisor can be manually invoked with the AUTOCONFIGURE

command from the command line processor (CLP), either standalone or as par
of the CREATE DATABASE command. Additionally, it can also be run via a GUI
available in the Control Center, or by calling the db2AutoConfig API, and finally
by using the ADMIN_CMD stored procedure.

To invoke this wizard from the DB2 Control Center, expand the object tree until
you find the database that you want to tune. Select the icon for the database,
right-click and select Configuration Advisor. Through several dialog windows,
the wizard collects information about the percentage of memory dedicated to
DB2, type of workload, number of statements per transaction, transaction
throughput, trade-off between recovery and database performance, number of
applications, and isolation level of applications connected to the database.
Based on the supplied answers, the wizard proposes configuration changes and
gives the option to apply the recommendations or save them as a task for the
Task Center for later execution as shown in Figure 10-11 on page 372. The result
window is presented in Figure 10-12 on page 373.

Chapter 10. Testing and tuning 371

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Figure 10-11 Scheduling Configuration Advisor

372 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

Figure 10-12 Configuration Advisor recommendations

Initial configuration recommendations can also be acquired through the text

based AUTOCONFIGURE command. Example 10-36 shows the sample
execution of the command.

Example 10-36 Sample AUTOCONFIGURE command

db2 AUTOCONFIGURE USING mem_percent 40 tpm 300 num_local_apps 80 isolation CS

apply none

[...]

Current and Recommended Values for Database Configuration

Description Parameter Current Value Recommended Value

---------------------------------------------------------------------------------------
Default application heap (4KB) (APPLHEAPSZ) = 256 256
Catalog cache size (4KB) (CATALOGCACHE_SZ) = 33 91
Changed pages threshold (CHNGPGS_THRESH) = 80 80
Database heap (4KB) (DBHEAP) = 1200 3552
Degree of parallelism (DFT_DEGREE) = 1 1
Default tablespace extentsize (pages) (DFT_EXTENT_SZ) = 32 32

Chapter 10. Testing and tuning 373

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

[...]

Table 10-5 lists all AUTOCONFIGURE command parameters.

Table 10-5 Parameters for the autoconfigure command

Keyword Valid values Default Explanation
value

mem_percent 1–100 25 Percentage of memory to dedicate to

DB2

Workload_type simple, mixed, mixed Type of workload: simple for

complex transaction processing, complex for
warehousing

num_stmts 1–1 000 000 10 Number of statements per unit of work

Tpm 1–200 000 60 Transactions per minute

admin_priority performance, both Optimize for better performance or

recovery, both better recovery time
is_populated yes, no yes Is the database populated with data?

num_local_apps 0–5 000 0 Number of connected local

Applications
num_remote_apps 0–5 000 10 Number of connected remote
Applications

Isolation RR, RS, CS, UR RR Isolation levels: Repeatable Read,

Read Stability, Cursor Stability,
Uncommitted Read

bp_resizeable yes, no yes Are buffer pools re-sizeable?

10.5.7 Design Advisor

Well designed indexes are essential to database performance. DB2 comes with
the Index Advisor utility, which recommends indexes for specific SQL queries.
Index Advisor can be invoked either using the db2advis command or using the
Design Advisor wizard from the Command Center or Control Center. This utility
accepts one or more SQL statements and their relative frequency, known as a
workload.

Several benefits of the Index Advisor are:

򐂰 Finding the best indexes for a problem query
򐂰 Finding the best indexes for a specified workload. When specifying the
workload, you can use the frequency parameter to prioritize the queries. You
can also limit disk space for the target indexes.

374 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

򐂰 Testing an index on a workload without having to create the index

To execute the index advisor against a specific database, we first must specify
the workload that will be run against the database. To do this from the command
line, we create a file that defines the workload. Example 10-37 shows the queries
that will be ran against the SAMPLE database.

Example 10-37 db2advis.in input file

--#SET FREQUENCY 100
SELECT COUNT(*) FROM EMPLOYEE;
SELECT * FROM EMPLOYEE WHERE LASTNAME='HAAS';
--#SET FREQUENCY 1
SELECT AVG(BONUS), AVG(SALARY) FROM EMPLOYEE
GROUP BY WORKDEPT ORDER BY WORKDEPT;
--#SET FREQUENCY 50
select FIRSTNME, lastname, deptname from department d, employee e where
d.deptno = e.workdept and e.lastName like 'W%'

We can then run the db2advis command and specify the db2advis.in as the
workload input script. Example 10-38 shows the syntax and output to execute the
index advisor; for more options run db2advis -h from the command line.

Example 10-38 Finding indexes for a particular query

db2inst1@db2server:~/ > db2advis -d sample -i db2advis.in -t 5

Using user id as default schema name. Use -n option to specify schema

execution started at timestamp 2009-09-18-02.38.35.785903
found [3] SQL statements from the input file
Recommending indexes...
total disk space needed for initial set [ 0.000] MB
total disk space constrained to [ 29.459] MB
Trying variations of the solution set.
0 indexes in current solution
[813.2663] timerons (without recommendations)
[813.2663] timerons (with current solution)
[0.00%] improvement
--
--
-- LIST OF RECOMMENDED INDEXES
-- ===========================
-- no indexes are recommended for this workload.
--
--
-- RECOMMENDED EXISTING INDEXES
-- ============================
-- RUNSTATS ON TABLE "DB2INST1"."EMPLOYEE" FOR SAMPLED DETAILED INDEX "DB2INST1"."XEMP2"
;
-- COMMIT WORK ;
--

Chapter 10. Testing and tuning 375

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

--
-- UNUSED EXISTING INDEXES
-- ============================
-- DROP INDEX "DB2INST1"."XEMP2";
-- ===========================
--
-- ====ADVISOR DETAILED XML OUTPUT=============
-- ==(Benefits do not include clustering recommendations)==
--
--<?xml version="1.0"?>
--<design-advisor>
--<statement>
--<statementnum>0</statementnum>
--<statementtext>
-- SELECT COUNT(*) FROM EMPLOYEE
--</statementtext>
--<objects>
--<identifier>
--<name>EMPLOYEE</name>
--<schema>DB2INST1</schema>
--</identifier>
--<identifier>
--<name>XEMP2</name>
--<schema>DB2INST1</schema>
--</identifier>
--</objects>
--<benefit>0.000000</benefit>
--<frequency>100</frequency>
--</statement>
--<statement>
--<statementnum>1</statementnum>
--<statementtext>
-- SELECT * FROM EMPLOYEE WHERE LASTNAME='HAAS'
--</statementtext>
--<objects>
--<identifier>
--<name>EMPLOYEE</name>
--<schema>DB2INST1</schema>
--</identifier>
--</objects>
--<benefit>0.000000</benefit>
--<frequency>100</frequency>
--</statement>
--<statement>
--<statementnum>2</statementnum>
--<statementtext>
-- SELECT AVG(BONUS), AVG(SALARY)
-- FROM EMPLOYEE GROUP BY WORKDEPT ORDER BY
-- WORKDEPT
--</statementtext>
--<objects>
--<identifier>
--<name>EMPLOYEE</name>
--<schema>DB2INST1</schema>

376 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch10.fm

--</identifier>
--<identifier>
--<name>XEMP2</name>
--<schema>DB2INST1</schema>
--</identifier>
--</objects>
--<benefit>0.000000</benefit>
--<frequency>1</frequency>
--</statement>
--</design-advisor>
-- ====ADVISOR DETAILED XML OUTPUT=============
--

27 solutions were evaluated by the advisor

DB2 Workload Performance Advisor tool is finished.

Launching the Index Advisor in a GUI environment

The Index Advisor can also be invoked as a GUI tool. From the Control Center,
expand the object tree to find the Database folder and right-click on the desired
database to select Design Advisor. The wizard guides you through all the
necessary steps, and also helps to construct a workload by looking for recently
executed SQL and XQuery queries, or looks through the recently used
packages. In order to get accurate recommendations, it is important to have the
current catalog statistics. With the Design Advisor there is an option to collect the
required basic statistics. However, this increases the total calculation time.
Figure 10-13 presents a sample Design Advisor window.

Chapter 10. Testing and tuning 377

7093ch10.fm Draft Document for Review October 16, 2009 10:21 am

Figure 10-13 The Design Advisor

378 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

11

Chapter 11. Advanced DB2 features

In this chapter we dive into the advanced features offered by DB2. During
conversion of your database from MySQL to DB2, consider the DB2 features and
functions explained within this chapter to enhance your application.

The features covered in this chapter are:

򐂰 XML
򐂰 Compression
򐂰 Partitioning
򐂰 Multidimensional clustering (MDC)
򐂰 Materialized query tables (MQT)
򐂰 User defined data types (UDT)

© Copyright IBM Corp. 2009. All rights reserved. 379

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

11.1 pureXML
DB2 pureXML allows native storage of XML data in a pre-parsed tree format
within a database table. More specifically, this is done using a special XML
column data type. However, unlike other databases on the market, DB2 does not
store XML data simply as character strings or CLOB or shred it into relational
data. When inserting XML data into a DB2 database, the data gets parsed and
fragmented with node-level granularity preserving the tree structures of the
original XML data. Moreover, during this process, the user can enable DB2 to
validate XML data against an XML schema registered with the database and
thus making sure that the data is always in the format their applications require.

DB2 pureXML provides superior performance when it comes to managing and

manipulating XML data. This is because XML documents do not need to be
parsed at query runtime and the advance indexing technology available with
XML in DB2 speeds the searches across and within the already parsed
documents.

DB2 pureXML is fully integrated with relational processing allowing you to

seamlessly query various types of data at once using SQL, XML or SQL/XML
mixed query languages. And unlike other database vendors, there is no internal
translation of XQuery into SQL.

Figure 11-1 on page 381 illustrates a simple example of how relational and XML
data are integrated within a single table. As shown in the figure, when creating a
table within a DB2 database, you can specify both relational and XML data types
in separate column definitions. This creates a table, where for every row of
relational data, there will be an XML document associated with that row.

Since it is possible that large size of XML documents will be stored, they are
physically stored in separate objects by default - these objects are called the
XML Data Area (XDA) objects. For most application scenarios, this provides
excellent performance. However, you do have the option of storing the XML
documents in the same physical space as the relational data by using the SET
INLINE LENGTH clause in the CREATE or ALTER statement. For more
information refer to the IBM DB2 Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

380 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

Figure 11-1 SQL and XML data integration

In order to query and update SQL and XML data, SQL and XQuery statements
can be used (SQL/XML: ISO standard ISO/IEC 9075-14:2003). Several
operations are available to directly modify not only full documents, but also parts
or sub-trees of XML documents without having to read, modify and reinsert them.
Using the XQuery language you can directly modify single values and nodes
within the XML document. XQuery is a fairly new, standardized query language
supporting path-based expressions. More information to XQuery can be found
here:
http://www.w3.org/TR/2007/REC-xquery-20070123/.

With pureXML, applications are not only able to combine statements from both
languages to query SQL and XML data; you can express many queries in plain
XQuery, in SQL/XML, or XQuery with embedded SQL. In certain cases, one of
the options to express your query logic may be more intuitive than another. In
general, the “right” approach for querying XML data needs to be chosen on a
case-by-case basis, taking the application's requirements and characteristics
into account. Example 11-1 shows a simple XQuery command.

Example 11-1 XQuery command

db2 => XQUERY db2-fn:xmlcolumn('CUSTOMER.INFO')/customerinfo/name

1
--------------------------------------------

Chapter 11. Advanced DB2 features 381

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

<name>Kathy Smith</name>
<name>Kathy Smith</name>
<name>Jim Noodle</name>
<name>Robert Shoemaker</name>
<name>Matt Foreman</name>
<name>Larry Menard</name>

6 record(s) selected.

Example 11-2 illustrates in a table format, a combined SQL/XML statement used

to retrieve the XML data.

Example 11-2 SQL/XQuery command

db2 => SELECT T.* FROM CUSTOMER C, XMLTABLE('$cu/customerinfo' PASSING C.INFO
as "cu" COLUMNS "NAME" VARCHAR (30) PATH 'name',"STREET" VARCHAR (30) PATH
'addr/street', "CITY" VARCHAR (30) PATH 'addr/city') AS T

NAME STREET CITY

------------------------------ ------------------------------ ------------
Kathy Smith 5 Rosewood Toronto
Kathy Smith 25 EastCreek Markham
Jim Noodle 25 EastCreek Markham
Robert Shoemaker 1596 Baseline Aurora
Matt Foreman 1596 Baseline Toronto
Larry Menard 223 NatureValley Road Toronto

6 record(s) selected.

Example 11-3 shows how to use a combined XQuery / SQL statement to retrieve
XML data based on some relational data.

Example 11-3 XQuery with embedded SQL

db2 => XQUERY db2-fn:sqlquery("SELECT INFO FROM CUSTOMER WHERE
CID=1001")/customerinfo/name

1
<name>Kathy Smith</name>

1 record(s) selected.

DB2 pureXML technology

Features of DB2 pureXML technology include:
򐂰 Reduce development time (no XML parsing anymore, validation and schema
repository building)

382 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

򐂰 pureXML data type and storage techniques for efficient management of

hierarchical structures common in XML documents.
򐂰 pureXML indexing technology to speed searches of subsets of XML
documents.
򐂰 New query language support (for XQuery and SQL/XML) based on industry
standards and new query optimization techniques.
򐂰 Industry-leading support for managing, validating, and evolving XML
schemes.
򐂰 Comprehensive administrative capabilities, including extensions to popular
database utilities.
򐂰 Integration with popular application programming interfaces (APIs) and
development environments.
򐂰 XML shredding and publishing facilities for working with existing relational
models.
򐂰 Enterprise proven reliability, availability, scalability, performance, security,
and maturity that you have come to expect from DB2

Benefits of DB2 pureXML technology

The benefits of DB2 pureXML technology include:
򐂰 Reduce development time and costs (no XML parsing anymore, validation
and schema repository building)
򐂰 Avoid breaking down XML to relational data (directly query and update XML)
򐂰 Increase agility through versatile XML schema evolution
򐂰 Improve insight by exploiting previously unmanaged XML data and quicker
query processing through XML-optimized storage and indexing
򐂰 Improved application performance
򐂰 Simplified operating environment
򐂰 Lower storage costs

11.2 Data compression

Data compression is a feature in DB2, which compresses row data to reduce
storage requirements, improve I/O efficiency, and provide quicker data access
from the disk. By enabling data compression with DB2, storage space can be
saved up to 80%, and performance could be greatly increased in database
systems with high disk I/O activity. Considering the fact that disk storage can
often be the most expensive component of a database system, this feature can

Chapter 11. Advanced DB2 features 383

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

result in tremendous cost savings. These savings will also extend to backup disk
space and more.

Two forms of data compression are currently available to you:

򐂰 Value compression involves removing duplicate entries for a value, storing
only one copy, and keeping track of the location of any references to the
stored value.
򐂰 Row compression involves replacing repeating patterns that span multiple
column values within a row with shorter symbol strings. The row compression
logic scans a table that is to be compressed for repetitive and duplicate data.
A compression dictionary contains short, numeric keys to that data, and in a
compressed row, these keys replace the actual data. We are discussing this
type of compression in this section.

Before you consider turning on row compression, you can inspect your tables to
see what potential savings to expect. Compressing and decompressing data is
effortless.

To enable row compression, you can use the COMPRESS YES keywords in
either the CREATE or ALTER TABLE statement, as shown in Example 11-4.

Example 11-4 Enabling row compression

db2inst1@db2server:~> db2 ALTER TABLE admin.owners COMPRESS YES
DB20000I The SQL command completed successfully.

After enabling row compression and loading or inserting 1-2 MB of data, DB2
automatically creates the compression dictionary. This dictionary contains the
frequently occurring patterns with an associated shorter 12-bit symbol. These
symbols then are used to replace the original data.

On heavily CPU-utilized systems, the compression and decompression may

cause overhead and decrease of performance. In cases, where the system is
mainly I/O bound, less pages must be read from disk, which results in a
performance gain. Data compression can also reduce the amount of buffer pool
memory needed, since pages fetched from disk stay compressed in main
memory while the page is actively used.

Figure 11-2 illustrates the mapping of repeating patterns in two table rows to
dictionary symbols representing those patterns. The end result is a compressed
data record shorter in length than the original uncompressed record - depicted by
the yellow rectangles representing the rows beneath the table.

384 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

Figure 11-2 Row compression

As of DB2 9.7, data compression had been extended to include all temporary
tables. Data compression for temporary tables reduces the amount of temporary
disk space that is required for large and complex queries increasing query
performance.

Index objects and indexes on compressed temporary tables can also be

compressed to reduce storage costs. This is especially useful for large online
transaction processing (OLTP) and data warehousing environments, where it is
common to have many large indexes. In both cases, index compression can
cause significant performance improvements in I/O-bound environments, and
little or no performance decrements in CPU-bound systems.

If compression is enabled on a table with an XML column, the XML data that is
stored in the XDA object is also compressed. A separate compression dictionary
for the XML data is stored in the XDA object. XDA compression is not supported
for tables whose XML columns were created prior to this version; for such tables,
only the data object is compressed.

11.3 Partitioning features

In this section, we introduce DB2 partitioning features.

11.3.1 Database partitioning feature

If you require increased processing power or scalability of your data beyond the
capabilities of a single server, you can achieve this with database partitioning.
Database Partitioning Feature (DPF) allows partitioning of a database across
multiple physical servers or within a large SMP server, as shown in Figure 11-3.
DPF brings about scalability, since you can add new servers and spread your
database across them. This of course means more CPU, more memory, more

Chapter 11. Advanced DB2 features 385

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

disk and is ideal for larger databases used for data warehousing, data mining,
online analytical processing or working with online transaction processing
workloads.

Figure 11-3 Database partitioning feature

The database partitioning feature uses a shared-nothing architecture. All data

can be distributed and split across the participating nodes. From a client
perspective, when a DB2 database is split across different nodes using DPF,
applications and administrators will still see a single database copy and access
data within it as if it was never partitioned. When scanning through data, DB2 will
automatically parallelize queries by splitting them up and sending them directly to
the nodes that hold the parts of the data requested.

Such systems can also be made highly available when using shared storages. In
this case two or more nodes can share the file systems holding the table spaces.
If an outage occurs, the surviving node can immediately access the failed node's
table spaces and continue processing.

Keep in mind that database partitioning can be efficiently used in combination

with table partitioning and multidimensional clustering.

386 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

11.3.2 Table partitioning

Table partitioning allows for organization of table data into multiple storage
objects called data partitions. The table can be organized based on one or more
column values. The main benefit of table partitioning is that the partitions can be
distributed on to multiple table spaces and can greatly increase table capacity
limit. While the data externally appears as a single table, internally DB2 can
increase query performance by optimizing the access plans to use the method of
partition elimination. Most importantly table partitioning makes managing table
data easier using roll-in and roll-out operations.

To partition a particular table, specify the PARTITION BY RANGE clause and the
partitioning column(s). You can specify multiple columns and generated columns.
The column must be a base type, no LOBS, LONG VARCHARS, etc. Figure 11-4
on page 387 describes different syntax for creating the same table partitions. The
first CREATE table statement, creates a table with three partitions on the c1
column. This statement creates a partition to hold data for each of the following
ranges: 1 to 33, 34 to 66 and 67 to 99. We refer to this as short form because it
allows DB2 to create, name and distribute the partitions over three table spaces.
In the second CREATE TABLE statement, the user specifies the partition names
by using the PARTITION or PART key word. In this example the user also
specifies the table spaces, where each partition should be stored in.

Figure 11-4 Syntax for creating a Range Partitioned table

Chapter 11. Advanced DB2 features 387

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

After creating a partitioned table, open INSERT, UPDATE or LOAD into the
table, DB2 automatically inserts rows into the appropriate table partition
according to the specified range. If the inserted data does not fit within the
ranges of any of the partitions, DB2 produces an error.

Traditionally, in order to archive older data, data would need to be moved to the
archived locations and delete statements would be issued to remove the data
from the current table. This results in a full table scan to find all rows belonging to
the requested range. By using table partitioning, each table partition can be
quickly separated from the table using the DETACH PARTITION key words in
the ALTER TABLE statement. Example 11-5 describes the syntax for dropping a
particular table partition.

Example 11-5 Detaching a table partition

db2> ALTER TABLE sales DETACH PARTITION Q4_2009 INTO TABLE OldMonthSales

db2> COMMIT

db2> EXPORT OldMonthSales; DROP OldMonthSales

Physically, there is no impact to the system when using ALTER table DETACH
PARTITION command. The command is extremely fast, since no data
movement takes place. As you can see in Figure 11-5, the table containing the
detached partition resides in the same table space as the original table partition.
The DETACH only changes catalog entries to let DB2 know that the table
partition is no longer apart of the main table. Once the statement is committed,
the detached data is available from the new table name. From now on, the table
are a regular table and you can do whatever you need to it.

Figure 11-5 Detaching a table partition

The ATTACH command is very similar to DETACH. For more details visit the
IBM DB2 Information Center at:
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp

388 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

Another advantage to table partitioning is table partition elimination. This is

possible because DB2 uses the partition definition to see which partitions contain
the values requested. DB2 only queries those partitions with matching ranges.
Figure 11-6 on page 389 illustrates how DB2 uses query definitions for partition
elimination.

Figure 11-6 Example of a Range Partitioned table

Keep in mind, that table partitioning can be used efficiently in combination with
database partitioning and multidimensional clustering.

11.3.3 Multidimensional clustering

Multidimensional Clustering (MDC) provides a powerful method for improving
the performance of SELECT, INSERT, UPDATE, and DELETE statements.
Multidimensional clustering provides a way to organize data physically on disk
according to provided dimensions. The data stays clustered on disk, which can
significantly improve query performance and maintenance operations
(table/index reorganization). This feature is best-suited for data warehousing,
OLTP and large database environments.

The indexes for each dimension are block based, not record based, thus
reducing their size (and effort needed for logging and maintaining) dramatically.
Reorganization of the table in order to re-cluster is not necessary.

Example 11-6 shows the CREATE TABLE statement of an MDC table clustered
in three columns: itemId, nation, and orderDate. The block indexes for each
dimension are created automatically.

Example 11-6 Creating an MDC table

CREATE TABLE Orders(
itemId INT, nation varchar(30), orderDate date,
A INT, B INT, C Date, D INT …)
IN Tablespace X, Tablespace Y, Tablespace Z …
INDEX IN Tablespace Y

Chapter 11. Advanced DB2 features 389

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

ORGANIZE BY DIMENSIONS (itemId, nation, orderDate)

Figure 11-7 shows the data clustering according to three dimensions as defined
in Example 11-6.

Figure 11-7 MDC table and indexes

When organizing by dimensions, one or more table columns can be specified.

DB2 places all inserted rows with the same values for specific columns into a
physical location close to one another. This special physical location is called a
block. A block (extent) is a set of contiguous pages on disk, so access to these
records is sequential and accessed with minimal I/O operations.

If an existing block is filled, a new block is allocated. All blocks with the same
combination of dimension values are grouped into what is called cells.

With this internal organization DB2 can very quickly find data along dimensions
or find all rows for a specific combination of dimension values.

Benefits of MDC
Advantages of MDC tables:
򐂰 Dimension block index lookups can identify the required portions of the table
and quickly scan only the required blocks.
򐂰 Block indexes are smaller than record identifier (RID) indexes, thus lookups
are faster.

390 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

򐂰 Index ANDing and ORing can be performed at the block level and combined
with RIDs.
򐂰 Data is guaranteed to be clustered on extents, which makes retrieval faster.
򐂰 Rows can be deleted faster when rollout can be used.

Note that MDC can be efficiently used in combination with database partitioning
and table partitioning.

11.4 Materialized query tables (MQT)

A materialized query table (MQT) is a table whose definition is based on the
result of a query, and whose data is in the form of pre-computed results that are
taken from one or more tables on which the materialized query table definition is
based. The DB2 optimizer can use these tables to determine whether a query
can best be served by accessing an MQT instead of base tables.

Materialized query tables (MQTs) are a powerful way to improve response time
for complex queries, especially queries that might require some of the following
operations:
򐂰 Aggregated data over one or more dimensions
򐂰 Joins and aggregates data over a group of tables
򐂰 Data from a commonly accessed subset of data, that is, from a “hot”
horizontal or vertical database partition
򐂰 Repartitioned data from a table, or part of a table, in a partitioned database
environment

Knowledge of MQTs is integrated into the SQL and XQuery compiler. During
compilation, the query rewrites phases and the optimizer matches queries with
MQTs to determine whether substitution of an MQT for a query that accesses the
base tables is required. If an MQT is used, the EXPLAIN facility can provide
information about which MQT was selected.

Since MQTs behave like regular tables in many ways, the same guidelines as
MQTs apply for optimizing data access using table space definitions, creating
indexes, and issuing RUNSTATS.

As an example, assume a database warehouse contains a set of customers and

a set of credit card accounts. The warehouse records the set of transactions that
are made with the credit cards. Each transaction contains a set of items that are
purchased together. This schema is classified as a multi-star because it has two

Chapter 11. Advanced DB2 features 391

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

large tables, one containing transaction items and the other identifying the
purchase transactions.

We use three hierarchical dimensions to describe a transaction in our example:

product, location, and time. The product hierarchy is stored in two normalized
tables representing the product group and the product line. The location
hierarchy contains city, state, and country or region information and is
represented in a single de-normalized table. The time hierarchy contains day,
month, and year information and is encoded in a single date field. The date
dimensions are extracted from the date field of the transaction using built-in
functions. Other tables in this schema represent account information for
customers and customer information.

An MQT is created with the sum and count of sales for each level of the following
hierarchies:
򐂰 Product
򐂰 Location
򐂰 Time, composed of year, month, day.

Many queries can be satisfied from this stored aggregate data. The following
example shows how to create an MQT that computes sum and count of sales
along the product group and line dimensions; along the city, state, and country
dimension; and along the time dimension. It also includes several other columns
in its GROUP BY clause. Example 11-7 is an example of the create MQT
statement.

Example 11-7 create MQT statement

CREATE TABLE dba.PG_SALESSUM
AS (
SELECT l.id AS prodline, pg.id AS pgroup,
loc.country, loc.state, loc.city,
l.name AS linename, pg.name AS pgname,
YEAR(pdate) AS year, MONTH(pdate) AS month,
t.status,
SUM(ti.amount) AS amount,
COUNT(*) AS count
FROM cube.transitem AS ti, cube.trans AS t,
cube.loc AS loc, cube.pgroup AS pg,
cube.prodline AS l
WHERE ti.transid = t.id
AND ti.pgid = pg.id
AND pg.lineid = l.id
AND t.locid = loc.id
AND YEAR(pdate) > 1990
GROUP BY l.id, pg.id, loc.country, loc.state, loc.city,
year(pdate), month(pdate), t.status, l.name, pg.name

392 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

)
DATA INITIALLY DEFERRED REFRESH DEFERRED;

REFRESH TABLE dba.PG_SALESSUM;

The cost of computing the answer using the MQT could be significantly less than
using a large base table since a portion of the answer is already computed.
MQTs can reduce expensive joins, sorts, and aggregation of base data.

The larger the base tables become, the greater the improvements in response
time can be since the MQT grows more slowly than the base table. MQTs can
effectively eliminate overlapping work among queries by doing the computation
once the MQTs are built and refreshed, as well as reusing their content for many
queries.

11.5 User-defined data types (UDT)

In some cases existing built-in data types may not meet the needs of your
application. User-defined types (UDTs) allow you to create or extend existing
data types to your application needs.

There are six types of user-defined types:

򐂰 Distinct type
A distinct type is a user-defined data type that is based on an existing built-in
data type. Internally, it is stored as an existing data type, but it is considered
as a separate and incompatible type. The main advantages of using distinct
types are extensibility, strong typing, encapsulation, and customization.
A distinct type can be created by issuing the CREATE DISTINCT TYPE
statement. The following statement defines a new distinct type for our sample
application where we want all the identification numbers to have common
properties and functions. To achieve this goal we create a distinct type ID,
which contains INTERGER values:
db2> CREATE DISTINCT TYPE id AS integer with comparisons
򐂰 Structured type
A structured type is a user-defined data type that has a well defined structure
consisting of existing built-in or user-defined data types. A structured type has
attributes and methods defined. The attribute defines its data storage
properties and methods define its behavior.
A structured type may be used as the type of a table, view, or column. When
used as a type for a table or view, that table or view is known as a typed table

Chapter 11. Advanced DB2 features 393

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

or typed view, respectively. For typed tables and typed views, the names and
data types of the attributes of the structured type become the names and data
types of the columns of this typed table or typed view. Rows of the typed table
or typed view can be thought of as a representation of instances of the
structured type. When used as a data type for a column, the column contains
values of that structured type (or values of any of that type's subtypes, as
defined below). Methods are used to retrieve or manipulate attributes of a
structured column object.
A structured type can be created using the CREATE TYPE statement. For
example, we can define a product and sku type, which can be used to create
typed tables as shown in Example 11-8. Figure 11-8 shows its hierarchy.

Distinct type

ID

Structure type Typed table

Order use Product_type Product

Structure type Typed table

SKU_type SKU

Figure 11-8 User defined data types

Example 11-8 Structured types and typed tables

db2>CREATE TYPE product_type AS (
name VARCHAR(20), description VARCHAR(200), brand VARCHAR(30))
instantiable ref using integer mode db2sql

db2>CREATE TYPE sku_type

UNDER product_type AS (quantity integer)
instantiable mode db2sql

db2>CREATE TABLE product

OF product_type (ref is id user generated)

db2>CREATE TABLE sku of sku_type

UNDER product inherit select privileges

394 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ch11.fm

db2>CREATE TABLE order(id ID,sku sku_type)

We can also use this type as a type for a column as shown in the last
statement of Example 11-8 on page 394.
򐂰 Reference type
A reference type is a companion type to a structured type. Similar to a distinct
type, a reference type is a scalar type that shares a common representation
with one of the built-in data types. This same representation is shared for all
types in the type hierarchy. The reference type representation is defined
when the root type of a type hierarchy is created. When using a reference
type, a structured type is specified as a parameter of the type. This parameter
is called the target type of the reference.
The target of a reference is always a row in a typed table or a typed view.
When a reference type is used, it may have a scope defined. The scope
identifies a table (called the target table) or view (called the target view) that
contains the target row of a reference value. The target table or view must
have the same type as the target type of the reference type. An instance of a
scoped reference type uniquely identifies a row in a typed table or typed view,
called the target row
򐂰 Array Type
A user-defined array type is a data type that is defined as an array with
elements of another data type. Every ordinary array type has an index with
the data type of INTEGER and has a defined maximum cardinality. Every
associative array has an index with the data type of INTEGER or VARCHAR
and does not have a defined maximum cardinality.
򐂰 Row type
A row type is a data type that is defined as an ordered sequence of named
fields, each with an associated data type, which effectively represents a row.
A row type can be used as the data type for variables and parameters in SQL
PL to provide simple manipulation of a row of data.
򐂰 Cursor data type
A user-defined cursor type is a user-defined data type defined with the
keyword CURSOR and optionally with an associated row type. A user-defined
cursor type with an associated row type is a strongly-typed cursor type;
otherwise, it is a weakly-typed cursor type. A value of a user-defined cursor
type represents a reference to an underlying cursor.

Chapter 11. Advanced DB2 features 395

7093ch11.fm Draft Document for Review October 16, 2009 10:21 am

396 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ax01.fm

Appendix A. Mapping MySQL built-in

functions and operators
This appendix provides the full list of MySQL built-in functions and operators and
the DB2 equivalent function or solution. This section includes background and
examples for the built-in functions and operators.

© Copyright IBM Corp. 2009. All rights reserved. 397

7093ax01.fm Draft Document for Review October 16, 2009 10:22 am

A.1 Grouping related functions

Table A-1 lists the differences in two databases and provides conversion
examples.

Table A-1 MySQL and DB2 grouping related functions

MySQL function MYSQL example DB2 function DB2 example Notes

AVG([DISTINCT] mysql> SELECT a, AVG ([DISTINCT | ALL] db2 " SELECT a, Returns the
expression) AVG(b) expression) AVG(b) average set of
FROM t1 FROM t1 numbers
GROUP BY a GROUP BY a"

COUNT([DISTINCT] mysql> SELECT a, COUNT([DISTINCT| db2 " SELECT a, Returns the

expression, COUNT(b) ALL] expression). COUNT(b) number of rows or
expression,...) FROM t1 FROM t1 values in a set of
GROUP BY a DB2 allows only one GROUP BY a" rows or values.
expression:
Use CONCAT for
character data type or
CHAR and CONCAT on
numeric data types

MAX ([DISTINCT] mysql> SELECT a, MAX ([DISTINCT | ALL] db2 "SELECT a, Returns the
expression) MAX(b) expression) MAX(b) maximum value in
FROM t1 FROM t1 a set of values.
GROUP BY a GROUP BY a"

MIN ([DISTINCT] mysql> SELECT a, MIN ([DISTINCT | ALL] db2 "SELECT a, MIN(b) Returns the
expression) MIN(b) expression) FROM t1 minimum value in a
FROM t1 GROUP BY a" set of values.
GROUP BY a

STDDEV (expression) / mysql> SELECT STDDEV ([DISTINCT | db2 " SELECT Returns the
STDDEV_POP STDDEV (a),a ALL] expression) stddev(a),a standard deviation
(expression) FROM t1 FROM t1 (/n) of a set of
GROUP BY a GROUP BY a" numbers

SUM([DISTINCT] mysql> SELECT a, SUM([DISTINCT | ALL] db2 " SELECT a, Returns the sum of
expression) SUM(b) expression) sum(b) a set of numbers.
FROM t1 FROM t1
GROUP BY a GROUP BY a"

VAR_POP(expression) / mysql> SELECT VARIANCE ([DISTINCT | db2 " SELECT Returns the
VARIANCE(expression) VAR_POP(a) ALL] expression) VARIANCE(a) variance of a set of
FROM t1 FROM t1 numbers.
GROUP BY a GROUP BY a"

BIT_AND (expression) mysql> SELECT No equivalent function. Refer to UDF B.1, Returns the the
This function is an BIT_AND(a), a Implement using UDF. “Sample code for value of the bitwise
extension to SQL standards FROM t1 BIT_AND” on page 412 logical AND
GROUP BY a operation

BIT_OR (expression) mysql> SELECT No equivalent function. Refer to UDF B.1, Returns the the
This function is an BIT_OR(a), a Implement using UDF “Sample code for value of the bitwise
extension to SQL standards FROM t1 BIT_AND” on page 412 logical OR
GROUP BY a operation

BIT_XOR (expression) mysql> SELECT No equivalent function. Refer to UDF B.1, Returns the the
This function is an BIT_XOR(a), a Implement using UDF “Sample code for value of the bitwise
extension to SQL standards FROM t1 BIT_AND” on page 412 logical XOR
GROUP BY a operation

398 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ax01.fm

GROUP_CONCAT(expr mysql> SELECT a, No equivalent function. Returns a

ession) GROUP_CONCAT(b) Implement using UDF concatenated
This function is an FROM t1 group by a variable for all
extension to SQL values.
standards

STD mysql> SELECT STDDEV ([DISTINCT | db2 "SELECT Returns the

This function is an STD(a),a ALL] expression) stddev(a), a standard deviation
extension to SQL standards FROM t1 FROM t1 (/n) of a set of
GROUP BY a GROUP BY a" numbers

Not Available CORRELATION db2 " SELECT Returns the

(expression, expression) CORRELATION(a, b) coefficient of
FROM t1 correlation of a set
WHERE c = 52“ of number pairs.

Not Available COVARIANCE db2 “SELECT Returns the

(expression, expression) COVARIANCE(a, b) (population)
FROM t1 covariance of a set
WHERE c = 52 “ of number pairs.

Not Available GROUPING(expression) db2 " SELECT a, b, Returns a value

SUM(c) AS x, that indicates
GROUPING(a) AS whether or not a
y, row returned in a
GROUPING(b) AS z GROUP BY answer
FROM t1 set is a row
GROUP BY CUBE generated by a
(a, b) grouping set that
ORDER BY a, b” excludes the
column
represented by
expression

GROUP BY on alias mysql> SELECT a as x use column name for db2 " SELECT a
FROM a grouping FROM t1
GROUP BY x; GROUP BY a"

GROUP BY on position mysql> SELECT a use column name for db2 " SELECT a
FROM t1 grouping FROM t1
GROUP BY 1 GROUP BY a"

HAVING on alias mysql> SELECT a as x use column name in db2 " SELECT a
FROM t1 having clause FROM t1
GROUP BY a GROUP BY a
HAVING x > 0 HAVING a > 0”

Appendix A. Mapping MySQL built-in functions and operators 399

7093ax01.fm Draft Document for Review October 16, 2009 10:22 am

A.2 String functions

Table A-2 provides an overview of MySQL string related functions, and how
these can be converted to DB2.

Table A-2 MySQL and DB2 string functions

MySQL MYSQL example DB2 function DB2 example Notes
function

ASCII(string) mysql> SELECT ascii('a'); ASCII(string) db2 "VALUES ascii('a') " Returns ASCII
+------------+ code value
| ascii('a') | 1
+------------+ -----------
| 97 | 97
+------------+
1 row in set (0.00 sec) 1 record(s) selected

BIN(integer) mysql> SELECT bin(5); No equivalent Returns the binary

+--------+ function. value of the string
| bin(5) | Alternative use
+--------+ UDF
| 101 |
+--------+
1 row in set (0.00 sec)

BIT_LENGTH( mysql> SELECT bit_length('Hello') ; LENGTH(string, db2 "VALUES Result depends on

string) +---------------------+ CODEUNITS16| length('Hello')*8 " the
| bit_length('Hello') | CODEUNITS32| encoding scheme
+---------------------+ OCTETS) 1 used for
| 40 | ----------- character data:
+---------------------+ 40 singly byte,
1 row in set (0.00 sec) double byte,
1 record(s) selected. UTF-8

CHAR_LENGT mysql> SELECT CHARACTER_L db2 " VALUES

H(string) / CHAR_LENGTH('Orange'); ENGTH(string, CHAR_LENGTH('Orange', Returns the
CHARACTER_ +-----------------------+ CODEUNITS16| CODEUNITS32)" number of
LENGTH(string | CHAR_LENGTH('Orange') | CODEUNITS32| bytes for
) +-----------------------+ OCTETS),/ 1 expression. For
| 6| CHAR_LENGT ----------- double byte
+-----------------------+ H(string, 6 character set
1 row in set (0.00 sec) CODEUNITS16| (DBCS) the
CODEUNITS32| 1 record(s) selected. number of
OCTETS) DBCS characters
is
returned

CHAR(int, mysql> SELECT char(97); CHR(string) db2 "VALUES chr('97') " Returns the
[USING +----------+ character that has
character set]) | char(97) | 1 the ASCII code
+----------+ - value specified by
|a | a the argument
+----------+
1 row in set (0.00 sec) 1 record(s) selected.

400 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ax01.fm

CONCAT_WS( mysql> SELECT CONCAT_WS('-', Use || to db2 "SELECT (firstName || '-' || Returns the
separator, firstname, lastname, loginname ) as implement lastName || '-' || loginName) as concatenation of
string, FULLNAME from owners where id = CONCAT(list) fullName from admin.owners string arguments
string,…) 501; where id = 501" with separator
+-------------------------+
| FULLNAME | FULLNAME
+-------------------------+ -----------------------------------------
| Angela-Carlson-acarlson | ---------------------
+-------------------------+ Angela-Carlson-acarlson
1 row in set (0.01 sec)
1 record(s) selected.

CONCAT(strin mysql> SELECT Use db2 "SELECT (firstName || ' ' || Returns the
g, string,…) CONCAT(firstname, ' ', lastname) as CONCAT(string, lastName) as fullName from concatenation of
FULLNAME from owners where id = string) or || to admin.owners where id = 501" string arguments
501; implement
+----------------+ CONCAT(list) FULLNAME
| FULLNAME | -----------------------------------------
+----------------+ Angela Carlson
| Angela Carlson | 1 record(s) selected.
+----------------+
1 row in set (0.00 sec)

ELT(int,str1,str mysql> SELECT ELT(2, firstname, No equivalent Returns the

2,str3,...) lastname, loginname ) as function. Returns the nth
FULLNAME from owners where id = Alternative use string or
501; CASE NULL.
+----------+ expression or
| FULLNAME | UDF.
+----------+
| Carlson |
+----------+
1 row in set (0.00 sec)

EXPORT_SET( mysql> SELECT No equivalent Returns a string

bit, onString, EXPORT_SET(12,'ON','OFF','|',4); function. representation for
offString, +---------------------------------+ Alternative use each of the bit
length) | EXPORT_SET(12,'ON','OFF','|',4) | UDF values of a binary
+---------------------------------+ value.
| OFF|OFF|ON|ON |
+---------------------------------+
1 row in set (0.00 sec)

FIELD(string,st mysql> SELECT FIELD('Carlson', No equivalent Returns the

ring1,string2,str firstname, lastname, loginname ) as function. position of the
ing3,...)) FULLNAME from owners where id = Alternative use matching string.
501; CASE
+----------+ expression or
| FULLNAME | UDF
+----------+
| 2|
+----------+
1 row in set (0.00 sec)

FIND_IN_SET( mysql> SELECT No equivalent Returns the

expr) FIND_IN_SET('Doe','John,Doe,Stas function. position in the set
ie,Smith') as POSITION; Implement using of the matching
+----------+ UDF string.
| POSITION |
+----------+
| 2|
+----------+
1 row in set (0.01 sec))

Appendix A. Mapping MySQL built-in functions and operators 401

7093ax01.fm Draft Document for Review October 16, 2009 10:22 am

FORMAT(doub FORMAT: No equivalent Refer to UDF B.2, “Sample Returns the

le, integer) select function. code for FORMAT function” on rounded string
format(1234.5555, 2) Implement using page 413
returns 1,234.56 UDF

HEX(string) mysql> SELECT HEX(255); HEX(string) db2 "VALUES Returns a

+----------+ HEX(00000255)" hexadecimal
| HEX(255) | representation of a
+----------+ 1 value as a
| FF | -------- character string.
+----------+ FF000000
1 row in set (0.00 sec)
1 record(s) selected.

INSERT(string, mysql> SELECT INSERT('original', INSERT(string, db2 "VALUES Inserts a string

position, length, 2, 2, 'NEW'); position, length, INSERT('original', 2, 2, 'NEW')" into an existing
substring) +---------------------------------+ substring) 1 string
| INSERT('original', 2, 2, 'NEW') | -----------
+---------------------------------+ oNEWginal
| oNEWginal | 1 record(s) selected.
+---------------------------------+
1 row in set (0.00 sec)

INSTR mysql> SELECT LOCATE('N', LOCATE(substri db2 " SELECT LOCATE('N', Returns the
(substring, 'DINING') ng, string, [start], 'DINING') starting position of
string) -> ; [CODEUNITS16 FROM the first
/LOCATE(subs +-----------------------+ | SYSIBM.SYSDUMMY1" occurrence of one
tring, string, | LOCATE('N', 'DINING') | CODEUNITS32| 1 string within
[position]) +-----------------------+ OCTETS]) ----------- another string
/POSITION(su | 3| 3
bstring, string) +-----------------------+ 1 record(s) selected.
1 row in set (0.00 sec)

LCASE(string) / mysql> SELECT LCASE('JOB'); LCASE(string) db2 " SELECT LCASE('JOB') Returns a string in
LOWER(string) +--------------+ FROM which all
| LCASE('JOB') | SYSIBM.SYSDUMMY1" characters have
+--------------+ 1 been converted to
| job | --- lowercase
+--------------+ job characters
1 row in set (0.00 sec) 1 record(s) selected.

LEFT(string1, mysql> SELECT LEFT('Inventory', LEFT(string1, db2 "SELECT Returns a string

strin2) 6); integer) LEFT('Inventory', 6) FROM consisting of the
+----------------------+ SYSIBM.SYSDUMMY1" leftmost String up
| LEFT('Inventory', 6) | 1 to the integer
+----------------------+ --------- position
| Invent | Invent
+----------------------+ 1 record(s) selected.
1 row in set (0.00 sec)

LENGTH(string LENGTH(string) db2 "select LENGTH('Jürgen') Returns the length

)/ mysql> select LENGTH('Jürgen') ; / from admin.owners where id = of string in the
OCTECT_LEN +-------------------+ OCTET_LENGT 501" implicit or explicit
TH(string) | LENGTH('Jürgen') | H(string) 1 string unit.
+-------------------+ -----------
| 7| 7
+-------------------+ 1 record(s) selected.
1 row in set (0.00 sec)

LOAD_FILE(dir update blobTBL SET data = Use the LOAD Inserts the file into
String) LOAD_FILE('/tmp/AquaBlue.jpg') command with the database.
WHERE id = 6; LOBS FROM
<lob_directory>

402 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ax01.fm

LPAD(string, mysql> SELECT No equivalent Refer to UDF B.3, “Sample Returns the a
length, LPAD('TEST',6,'!!!'); function. code for RPAD and LPAD string of the given
substring) / +----------------------+ Implement functions” on page 414 length, if the length
RDAP( string, | LPAD('TEST',6,'!!!') | using UDF is longer then the
length, +----------------------+ string the
substring) | !!TEST | substring
+----------------------+ characters will be
1 row in set (0.00 sec) added to the left or
right end.

LTRIM(string) mysql> SELECT LTRIM(' Apple'); LTRIM(string) db2 "SELECT LTRIM(' Apple') Removes blanks
+------------------+ FROM from the beginning
| LTRIM(' Apple') | SYSIBM.SYSDUMMY1" of
+------------------+ 1 string-expression
| Apple | -------
+------------------+ Apple
1 row in set (0.00 sec)

MAKE_SET(bit mysql> SELECT MAKE_SET(2|4, No equivalent Returns a set of

s, string, "HELLO", "BYE", "GOOD DAY", function. strings based on
string,....) "GOOD NIGHT") as OUTPUT; Alternative use the bit.
+--------------+ UDF
| OUTPUT |
+--------------+
| BYE,GOOD DAY |
+--------------+
1 row in set (0.00 sec)

OCT(expressio mysql> select oct(22); No equivalent Returns the octal

n) +---------+ function. value of the
| oct(22) | Alternative use expression.
+---------+ UDF
| 26 |
+---------+

QUOTE(string) mysql> SELECT quote(firstname) SELECT with || db2 "select ('''' || firstname || '''') Returns string with
from owners where id = 501; from admin.owners where id = single quotes.
+------------------+ 501"
| quote(firstname) |
+------------------+ 1
| 'Angela' | ----------------------
+------------------+ 'Angela'
1 row in set (0.00 sec)
1 record(s) selected.

REPEAT(string mysql> select REPEAT('REPEAT REPEAT(string, db2 "VALUES Returns the

, integer) THIS ', 5); integer) REPEAT('REPEAT THIS ', 5)" repeated string N
+---------------------------------------------- times.
----------------+ 1
| REPEAT('REPEAT THIS ', 5) -------
| REPEAT THIS REPEAT THIS
+---------------------------------------------- REPEAT THIS REPEAT THIS
----------------+ REPEAT THIS
| REPEAT THIS REPEAT THIS 1 record(s) selected.
REPEAT THIS REPEAT THIS
REPEAT THIS |
+----------------------------------------------
----------------+
1 row in set (0.00 sec)

Appendix A. Mapping MySQL built-in functions and operators 403

7093ax01.fm Draft Document for Review October 16, 2009 10:22 am

REPLACE(strin mysql> SELECT REPLACE REPLACE(strin db2 "VALUES REPLACE Returns as sting
g1, string2, ('DINING', 'N', 'VID'); g1, string2, ('DINING', 'N', 'VID') " with all
string3) +--------------------------------+ string3) 1 occurrences of
| REPLACE ('DINING', 'N', 'VID') | ------------------ string2 in string1
+--------------------------------+ DIVIDIVIDG with string3
| DIVIDIVIDG | 1 record(s) selected.
+--------------------------------+
1 row in set (0.00 sec)

REVERSE(stri mysql> SELECT REVERSE('abc'); SET (RESTSTR, LEN) = Returns the

ng1, S) +----------------+ (INSTR, reverse order of
| REVERSE('abc') | LENGTH(INSTR)); the string.
+----------------+ WHILE LEN > 0 DO Implement UDF.
| cba | SET (REVSTR, RESTSTR, For the complete
+----------------+ LEN)= code of the
1 row in set (0.00 sec) (SUBSTR(RESTSTR, 1, 1) || example refer to
REVSTR,SUBSTR(RESTSTR, IBM DB2 SQL
2, Reference,
LEN - 1),LEN - 1); Volume 1, V8,
END WHILE; SC10-4249

RIGHT(string, mysql> SELECT RIGHT('Inventory', RIGHT(string, db2 "VALUES Returns a string

length) 4); length) RIGHT('Inventory', 4 ) " consisting of the
+-----------------------+ 1 rightmost String
| RIGHT('Inventory', 4) | --------- starting from the
+-----------------------+ tory integer position
| tory | 1 record(s) selected.
+-----------------------+
1 row in set (0.00 sec)

RTRIM(string) mysql> SELECT RTRIM('PEAR '); RTRIM(string) db2 "VALUES RTRIM('PEAR Removes blanks
+------------------+ ') " from the end of
| RTRIM('PEAR ') | string
+------------------+ 1
| PEAR | -------
+------------------+ PEAR
1 row in set (0.00 sec) 1 record(s) selected.

SOUNDEX(stri mysql> SELECT SOUNDEX(strin db2 "VALUES Returns a

ng) SOUNDEX('apple'); g) SOUNDEX('apple')" 4-character code
+------------------+ 1 representing the
| SOUNDEX('apple') | ---- sound of the words
+------------------+ A140 in the argument.
| A140 | 1 record(s) selected. MySQL: String1
+------------------+ SOUNDS LIKE
1 row in set (0.00 sec) string2 can be
ported to
SOUNDEX(String
1) =
SOUNDEX(string
2) in DB2

SPACE(expres mysql> SELECT space(30); SPACE(express db2 " VALUES space(3)" Returns a
sion) +--------------------------------+ ion) 1 character string
| space(30) | ----------- consisting of
+--------------------------------+ blanks with length
| | specified by the
+--------------------------------+ second argument.
1 row in set (0.00 sec)

404 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ax01.fm

STRCMP(strin mysql> SELECT STRCMP('test', CASE Implement using CASE Returns -1 if the
g, string) 'testing'); expression and VALUES first string is
+----------------------------+ statement smaller, 0 if the
| STRCMP('test', 'testing') | strings are the
+----------------------------+ same length, 1 if
| -1 | the second string
+----------------------------+ is smaller
1 row in set (0.00 sec)

SUBSTRING_I mysql> SELECT No equivalent Refer to UDF B.7, “Sample

NDEX() SUBSTRING_INDEX('This is a function. code for SUBSTRING_INDEX”
test...', 't', 2); Implement on page 430
+---------------------------------------------- using UDF
+
| SUBSTRING_INDEX('This is a
test...', 't', 2) |
+----------------------------------------------
+
| This is a tes |
+----------------------------------------------
+
1 row in set (0.00 sec)

SUBSTRING(st mysql> select substring('abcdef', 2, SUBSTR(string, db2 "VALUES('abcdef', 2, 3)” Returns a

ring, position, 3); position, length) 1 substring of a
length) / MID +---------------------------+ --- string.
(string, | substring('abcdef', 2, 3) | bcd
position, +---------------------------+
length) / | bcd | 1 record(s) selected.
SUBSTR(string +---------------------------+
, position, 1 row in set (0.00 sec)
[length])

TRIM([Both | mysql> select trim(trailing from TRIM([Both | db2 "VALUES trim(trailing from Removes blanks
Leading | trim(LEADING FROM ' abc ')) as Leading | trailing trim(LEADING FROM ' abc '))" or occurrences of
trailing OUTPUT; [substring] another specified
[substring] +--------+ FROM] string) OUTPUT character from the
FROM] string) | OUTPUT | --------- end or the
+--------+ abc beginning of a
| abc | string expression
+--------+ 1 record(s) selected.
1 row in set (0.00 sec)

UCASE(string) mysql> SELECT UPPER('jobs'); UCASE(string) / db2 "VALUES UPPER('jobs')" Returns a string in
/ UPPER +---------------+ UPPER (String) 1 which all
(String) | UPPER('jobs') | ---- characters have
+---------------+ JOBS been converted to
| JOBS | 1 record(s) selected. uppercase
+---------------+ characters
1 row in set (0.00 sec)

UNHEX() mysql> SELECT No equivalent Converts

UNHEX('546F646179'); function. hexadecimal digits
+---------------------+ Alternative use to a character
| UNHEX('546F646179') | UDF string.
+---------------------+
| Today |
+---------------------+
1 row in set (0.00 sec)

Appendix A. Mapping MySQL built-in functions and operators 405

7093ax01.fm Draft Document for Review October 16, 2009 10:22 am

A.3 Numeric functions

Table A-3 lists some functions specific to manipulating numbers.

Table A-3 Numeric functions

MySQL Command MySQL example DB2 command DB2 example

ABS(expression) mysql> SELECT ABS(expression) db2 "VALUES ABS(-51234)"

ABS(-51234);

ACOS(expression) mysql> SELECT COS(1); ACOS(expression) db2 "VALUES COS(1)"

ASIN(expression) mysql> SELECT ASIN(1); ASIN(expression) db2 "VALUES ASIN(1)"

ATAN(expression, mysql> SELECT ATAN(expression), db2 "VALUES ATAN2(-1,1)"

expression),ATAN2( ATAN2(1,-1); ATAN2(expression,
expression, expression), expression),

CEIL(expression), mysql> SELECT CEIL(expression), db2 "VALUES Ceiling(3.35)"

CEILING(expression) CEILING(3.35); CEILING(expression)

COS(expression) mysql> SELECT cos(0); COS(expression) db2 "VALUES COS(0)"

COT(expression) mysql> SELECT coT(1); COT(expression) db2 "VALUES COT(1)"

DEGREES(expression) mysql> SELECT DEGREES(expression) db2 "VALUES DEGREES(3)"

DEGREES(3);

EXP(expression) mysql> SELECT EXP(3); EXP(expression) db2 "VALUES EXP(3)"

FLOOR(expression) mysql> SELECT FLOOR(expression) db2 "VALUES FLOOR(3.35)".

FLOOR(3.35);

LN(expression) mysql> SELECT LN(expression) db2 "VALUES LN(100)"

LN(100);
LOG10(expression) mysql> SELECT LOG10(expression) db2 "VALUES LOG10(100)"
LOG10(100);

LOG(expression, mysql> SELECT LOG(expression) db2 "VALUES LOG(100)"

[expression]) LOG(100);

MOD(expression, mysql> SELECT MOD(expression, db2 "VALUES MOD(125, 50)"

expression) MOD(125, 50); expression)
POW(expression, mysql> SELECT POWER(expression, db2 "VALUES POWER(5, 2)"
expression), POWER(5, 2); expression)
POWER(expression,
expression)

RADIANS(expression) mysql> SELECT RADIANS(expression) db2 "VALUES

RADIANS(180); RADIANS(180)"

RAND([expression]) mysql> SELECT RAND(); RAND([expression]) db2 "VALUES RAND()"

ROUND(expression, mysql> SELECT ROUND(expression, db2 "VALUES

expression) ROUND(873.726, 1); expression) ROUND(873.726, 1)"

SIGN(expression) mysql> SELECT SIGN(expression) db2 "VALUES SIGN(6453)"

SIGN(6453);

406 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ax01.fm

MySQL Command MySQL example DB2 command DB2 example

SIN(expression) mysql> SELECT SIN(3); SIN(expression) db2 "VALUES SIN(3)"

SQRT(expression) mysql> SELECT SQRT(expression) db2 "VALUES SQRT(25)"

SQRT(25);

TAN(expression) mysql> SELECT TAN(2); TAN(expression) db2 "VALUES TAN(2)"

TRUNCATE(expression, mysql> SELECT TRUNCATE(expression, db2 "VALUES
expression) TRUNCATE(873.726,2); expression) TRUNCATE(873.726,2)"

A.4 Date and time functions

Table A-4 lists some date and time related functions.

Table A-4 Date and Time related functions

MySQL command MySQL Example DB2 command DB2 example

ADDDATE(), mysql> SELECT DATE + expression db2 "VALUES DATE('2009-8-31')

DATE_ADD(),DATE_SUB(), DATE_ADD('2009-8-31', + 15 days"
SUBDATE() INTERVAL 15 DAY);

CURDATE(), mysql> SELECT CURRENT DATE db2 "VALUES CURRENT DATE"

CURRENT_DATE(), CURDATE();
CURRENT_DATE

CURRENT_TIME(), mysql> SELECT CURRENT TIME db2 "VALUES CURRENT TIME”

CURRENT_TIME, CURRENT_TIME();
CURTIME()

DAYOFMONTH mysql> SELECT Day(expression) db2 "VALUES DAY('2009-08-31

(expression), DAYOFMONTH('2009-0 05:06:00')"
DAY(expression) 8-31 05:06:00');

EXTRACT (unit FROM mysql> SELECT Use concatenate different date db2 "VALUES (YEAR('2009-08-31
expression) EXTRACT(YEAR_MON stripping functions (DAY, YEAR, 05:06:00') || MONTH('2009-08-31
TH from '2009-08-31 MONTH, DAYNAME, DAYOF 05:06:00'))"
05:06:00'); WEEK, and so on.

HOUR(expression) mysql> SELECT MIDNIGHT_SECONDS db2 "VALUES

HOUR('2009-08-31 (expression)/60/60 MIDNIGHT_SECONDS('2009-08-
05:06:00'); 31 05:06:00')/60/60"

SEC_TO_TIME() UDF OR SELECT with time SEC_TO_TIME(arg INTEGER)

Arithmetic with HOUR, MINUTE, RETURNS TIME CONTAINS SQL
SECOND NO EXTERNAL ACTION
DETERMINISTIC RETURN
TIME('00:00:00') + (arg / 3600)
HOURS + MOD(arg / 60, 60)
MINUTES + MOD(arg, 3600)
SECONDS "

STR_TO_DATE(expression, mysql> SELECT DATE and TO_DATE db2 "VALUES

format) STR_TO_DATE('15,03,2 DATE(TO_DATE('15,3,2009','DD,
009','%d,%m,%Y'); MM,YYYY'))"

WEEK(expression) mysql> SELECT WEEK(expression) db2 "VALUES WEEK('2009-08-31

WEEKOFYEAR week('2009-08-31 05:06:00')"
(expression, mode) 05:06:00');

Appendix A. Mapping MySQL built-in functions and operators 407

7093ax01.fm Draft Document for Review October 16, 2009 10:22 am

MySQL command MySQL Example DB2 command DB2 example

WEEKDAY(expression) mysql> SELECT DAYOFWEEK(expression) db2 " VALUES

weekday('2009-08-31 DAYOFWEEK('2009-08-31
05:06:00'); 05:06:00')"
db2 " VALUES
DAYOFWEEK_ISO
('2009-08-31 05:06:00')"

A.5 Comparing operators and other functions

Table A-5 shows MySQL and DB2 operators.

Table A-5 MySQL and DB2 operator comparison

MySQL DB2 Comments

logical NOT as '!' in VALUES CASE WHEN Implement using CASE expression and VALUES
SELECT list 1!=1 THEN 0 ELSE 1 END statement

% MOD In MySQL % is a synonym for modulo

& (bitwise and) not available. Implement using UDF Refer to UDF B.1, “Sample code for BIT_AND” on
page 412

logical AND as '&&' in CASE Implement using CASE expression and VALUES
SELECT list statement

not equal, <> or != in SELECT CASE Implement using CASE Expression®

SELECT list:select 1<>1 WHEN 1<> 1
THEN x
ELSE y
END

Function = in SELECT list: CASE Implement using CASE expression and VALUES
select (1=1) statement

BETWEEN in SELECT CASE Implement using CASE

expression and VALUES
statement

<< and >> (bitwise shifts) no equivalent Implement using power function:
MySQL:
SELECT (x>>y)
SELECT(x<<y)
DB2 :
SELECT(x/power(2,y))
SELECT(x*power(2,y)):

BIT_COUNT no equivalent, implement using UDF Refer to UDF B.6, “Sample code for BIT_COUNT” on
page 429

ENCRYPT ENCRYPT DB2 requires encryption Password

FIELD CASE Implement using CASE expression and VALUES

statement

GREATEST FnGratst Refer to UDF B.4, “Sample code for GREATEST

function” on page 420

IF CASE Implement using CASE expression and VALUES

statement

408 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:22 am 7093ax01.fm

IN on numbers in SELECT CASE Implement using CASE expression and VALUES

statement

IN on strings in SELECT CASE Implement using CASE expression and VALUES

statement

LOCATE as INSTR LOCATE Arguments are swapped

INTERVAL CASE Implement using CASE expression and VALUES

statement

LAST_INSERT_ID IDENTITY_VAL_LOCAL DB2 returns NULL,

MySQL rreturns the most recently assigned value for an
identity column.

LEAST FnLeastN See UDF example in B.5, “Sample code for LEAST” on
page 425

LIKE in SELECT CASE with LIKE Implement using CASE expression and VALUES
statement

LIKE ESCAPE in SELECT CASE with LIKE and ESCAPE Implement using CASE expression and VALUES
statement

LOG(m,n) LOG(m)/LOG(n) To convert logarithm to an arbitrary base

NOT in SELECT CASE Implement using CASE expression and VALUES

statement

NOT BETWEEN in SELECT CASE Implement using CASE expression and VALUES
statement

NOT LIKE in SELECT CASE Implement using CASE expression and VALUES
statement

PASSWORD ENCRYPT For encryption

POW POWER Returns value of arg1 to the power of arg2

REGEXP in SELECT no equivalent see article on developerWorks® for work-around

http://www-106.ibm.com/developerworks/db2/librar
y/techarticle/0301stolze/0301stolze.html

VERSION db2level To retrieve installed version of DBMS

Appendix A. Mapping MySQL built-in functions and operators 409

7093ax01.fm Draft Document for Review October 16, 2009 10:22 am

410 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

Appendix B. Sample code for user

defined functions
This appendix provides sample code to implement various MySQL built-in
functions not provided by DB2. We would like to thank the respective authors of
the UDFs.

© Copyright IBM Corp. 2009. All rights reserved. 411

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

B.1 Sample code for BIT_AND

Example B-1 shows conversion code for MySQL BIT_AND function.

Example: B-1 User-defined function to map BIT_AND

--
-- DB2 UDF(User-Defined Function) Samples for conversion
--
-- 2001/08/29
--
-- Name of UDF: BIT_AND (N1 Integer, N2 Integer)
--
-- Used UDF: None
--
-- Description: Returns bit by bit and of both parameters.
--
-- Author: TOKUNAGA, Takashi
--
--------------------------------------------------------------------------
CREATE FUNCTION BITAND (N1 Integer, N2 Integer)
RETURNS Integer
SPECIFIC BITANDMySQL
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN
WITH
Repeat (S, M1, M2, Ans) AS
(Values (0, N1, N2, 0)
Union All
Select S+1, M1/2, M2/2, Ans+MOD(M1,2)*MOD(M2,2)*power(2,S)
From Repeat
Where M1 > 0
AND M2 > 0
AND S < 32
)
SELECT ANS
FROM Repeat
WHERE S = (SELECT MAX(S)
FROM Repeat)
;

Example B-2 shows the results of the BITAND user-defined function.

Example: B-2 Results for UDF BITAND

SQL0347W The recursive common table expression "DB2ADMIN.REPEAT"

may contain an infinite loop. SQLSTATE=01605

--------------------------------------------------------------------------
values bitand(10,8);
---------------------------------------------------

1
-----------
8

--------------------------------------------------------------------------
values bitand(14,3);
---------------------------------------------------

412 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

1
-----------
2

--------------------------------------------------------------------------
values bitand(1038,78);
---------------------------------------------------

1
-----------
14

B.2 Sample code for FORMAT function

This section provides UDF for FORMAT function. Example B-3 shows the code
for a user-defined function emulating FORMAT.

Example: B-3 FORMAT user-defined function

--
-- DB2 UDF(User-Defined Function) Samples for conversion
--
-- Created: 2004/02/29
--
-- Name of UDF: FORMAT (X Decimal(31,10), D Integer)
--
-- Used UDF: None
--
-- Description: Returns truncated to the precision specified by D and a "," for each 3 digits as a
separator.
--
-- Author: TOKUNAGA, Takashi
--

------------------------------ Command Entered ------------------------------

CREATE FUNCTION FORMAT (X Decimal(31,10), D Integer)
RETURNS VARCHAR(50)
LANGUAGE SQL
SPECIFIC FORMAT_MySQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
BEGIN ATOMIC
DECLARE XN DECIMAL(21,0);
DECLARE RetVal VARCHAR(50);
SET RetVal = SUBSTR(CHAR(MOD(ABS(X), 1)), 22, D+1);
SET XN = ABS(X);

Main_Loop:
WHILE XN > 0 DO
SET RetVal = SUBSTR(CHAR(MOD(XN,1000)),19,3) || RetVal;
SET XN = XN/1000;
IF XN > 0 THEN
SET RetVal = ',' || RetVal;
ELSE
LEAVE Main_Loop;
END IF;
END WHILE;

RETURN CASE WHEN X < 0 THEN '-' ELSE '' END

Appendix B. Sample code for user defined functions 413

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

|| TRANSLATE(LTRIM(TRANSLATE(RetVal,' ','0')),'0',' ');

END
!

Example B-4 shows the results of the converted FORMAT.

Example: B-4 Converted FORMAT UDF result

------------------------------ Command Entered ------------------------------
SELECT N
, FORMAT(N, 2)
, FORMAT(N, 0)
FROM (VALUES 12.34567, -12.34567, 120034.567, 123400123456789.) S(N)!
-----------------------------------------------------------------------------
--Return result

N 2 3
-------------------------- ----------------------- -----------------------
12.34567 12.34 12.
-12.34567 -12.34 -12.
120034.56700 120,034.56 120,034.
23400123456789.00000 123,400,123,456,789.00 123,400,123,456,789.

4 record(s) selected.

B.3 Sample code for RPAD and LPAD functions

This section provides UDF for LPD and RPAD functions. Example B-5 shows
code for a user-defined function emulating RPAD.

Example: B-5 CREATE FUNCTION RPAD and sample usage

-- DB2 UDF(User-Defined Function) Samples for conversion

--
-- 2001/08/27, 09/27, 11/06
--
-- Name of UDFs: RPAD (C1 VarChar(4000), N integer, C2 VarChar(4000))
-- RPAD (I1 Integer, N integer, C2 Varchar(4000))
-- LPAD (C1 VarChar(4000), N integer, C2 Varchar(4000))
-- LPAD (I1 Integer, N integer, C2 Varchar(4000))
--
-- Used UDF: None
--
-- Description: Add repeatedly C2 to the right(RPAD) or left(LPAD) of parameter 1 (C1 or I1)
-- and return N byte.
--
-- Author: TOKUNAGA, Takashi
--

--------------------------------------------------------------------------
CREATE FUNCTION RPAD (C1 VarChar(4000), N integer, C2 VarChar(4000))
RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC RPADBase
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN

414 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

substr(C1 ||
repeat(C2,((sign(N-length(C1))+1)/2)*(N-length(C1)+length(C2))/(length(C2)+1-sign(length(C2)))),1,N)
;

Example B-6 shows the results of the converted RPAD function.

Example: B-6 Usage of UDF RPAD

SELECT char(rpad('ABCDE',12,'*.'),20) FROM SYSIBM.SYSDUMMY1;

---------------------------------------------------
1
--------------------
ABCDE*.*.*.*

1 record(s) selected.
--------------------------------------------------------------------------
SELECT char(rpad('ABCDE',3,'*.'),20) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------
ABC

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(rpad('ABCDE',20,'') || 'X',50) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------------------------------------
ABCDE X
1 record(s) selected.

UDF RPAD with the third parameter omitted is shown in Example B-7.

Example: B-7 RPAD Omitting the third parameter

CREATE FUNCTION RPAD (C1 VarChar(4000), N integer)

RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC RPADVarCharParm2
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
RPAD(C1,N,' ')
;

Running the RPAD function gives you the results shown in Example B-8.

Example: B-8 Results of RPAD omitting the third parameter

SELECT char(rpad('ABCDE',15) || 'X',50) FROM SYSIBM.SYSDUMMY1;

---------------------------------------------------

1
--------------------------------------------------
ABCDE X

Appendix B. Sample code for user defined functions 415

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(rpad('ABCDE',3) || 'X',50) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------------------------------------
ABCX

1 record(s) selected.

Function RPAD allows a set of different input arguments. Example B-9 shows two
more RPAD UDFs.

Example: B-9 RPAD with first parameter as integer, 2, and 3 parameters

CREATE FUNCTION RPAD (I1 Integer, N integer, C2 Varchar(4000))

RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC RPADIntParm3
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
RPAD(rtrim(char(I1)),N,C2)
;

CREATE FUNCTION RPAD (I1 Integer, N integer)

RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC RPADIntParm2
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
RPAD(rtrim(char(I1)),N,' ')
;

And Example B-10 shows the results of the previous UDFs.

Example: B-10 Results of RPAD with first parameter as integer, 2, and 3 parameters

SELECT char(rpad(927,12,'*.'),50) FROM SYSIBM.SYSDUMMY1;

---------------------------------------------------

1
--------------------------------------------------
927*.*.*.*.*

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(rpad(927,12,'') || 'X',50) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

416 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

1
--------------------------------------------------
927 X

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(rpad(9021,3),20) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------
902

1 record(s) selected.

The counterpart for RPAD are the LPAD functions, which are shown in
Example B-11.

Example: B-11 LPAD: CREATE FUNCTION and sample usage

CREATE FUNCTION LPAD (C1 VarChar(4000), N integer, C2 VarChar(4000))

RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC LPADBase
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN N > length(C1) THEN
substr(repeat(C2,(N-length(C1)+length(C2))/(length(C2)+1-sign(length(C2)))),1,N-length(C1)) || C1
ELSE substr(C1,1,N)
END
;

Results of LPAD look like Example B-12.

Example: B-12 Results of LPAD: CREATE FUNCTION and sample usage

SELECT char(lpad('ABCDE',15,'*.'),50) FROM SYSIBM.SYSDUMMY1;

---------------------------------------------------

1
--------------------------------------------------
*.*.*.*.*.ABCDE

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(lpad('ABCDE',3,'*.'),50) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------------------------------------
ABC

Appendix B. Sample code for user defined functions 417

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(lpad('ABCDE',15,'') || 'X',50) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------------------------------------
ABCDEX

1 record(s) selected.

As RPAD allows LPAD a different number and data type for input arguments,
Example B-13 shows LPAD without the third parameter.

Example: B-13 LPAD: Omitting the third parameter

CREATE FUNCTION LPAD (C1 VarChar(4000), N integer)
RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC LPADParm2
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LPAD(C1,N,' ')
;

The results of Example B-13 should look like those in Example B-14.

Example: B-14 Result of LPAD: Omitting the third parameter

SELECT char(lpad('ABCDE',15),20) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------
ABCDE

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(lpad('ABCDE',3),20) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------
ABC

1 record(s) selected.

Two more LPAD UDFs with different characteristics are shown in Example B-15.

Example: B-15 LPAD: The first parameter is integer, 2, and 3 parameters

CREATE FUNCTION LPAD (I1 Integer, N integer, C2 Varchar(4000))

418 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC LPADIntParm3
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LPAD(rtrim(char(I1)),N,C2)
;

--------------------------------------------------------------------------
CREATE FUNCTION LPAD (I1 Integer, N integer)
RETURNS VARCHAR(4000)
LANGUAGE SQL
SPECIFIC LPADIntParm2
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LPAD(rtrim(char(I1)),N,' ')
;

The results are shown in Example B-16.

Example: B-16 Results of LPAD: The first parameter is integer, 2, and 3 parameter

SELECT char(lpad(9021,15,'*.'),50) FROM SYSIBM.SYSDUMMY1;

---------------------------------------------------

1
--------------------------------------------------
*.*.*.*.*.*9021

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(lpad(9021,15,''),50) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------------------------------------
9021

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(lpad(9021,3),20) FROM SYSIBM.SYSDUMMY1;
---------------------------------------------------

1
--------------------
902

1 record(s) selected.

Appendix B. Sample code for user defined functions 419

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

B.4 Sample code for GREATEST function

Example B-17 is a set of UDF examples emulating the behavior of MySQL
GREATEST function. The various UDFs accept input parameters in varchar and
from 2 to 10 input parameters.

Example: B-17 User-defined functions to map GREATEST

--
-- DB2 UDF(User-Defined Function) Samples for conversion
--
-- 2001/08/28, 08/29
--
-- Name of UDF: GREATEST (P1 VarChar(254), P2 VarChar(254), ...)
--
--
-- Used UDF: None
--
-- Description: Returns greatest value of list of data.
--
-- Author: TOKUNAGA, Takashi
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle2
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 >= P2 THEN P1
ELSE P2
END
;
---------------------------------------------------
--
-- GREATEST function with three parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle3
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 >= P2
THEN CASE
WHEN P1 >= P3 THEN P1
ELSE P3
END
ELSE CASE
WHEN P2 >= P3 THEN P2
ELSE P3
END
END
;
---------------------------------------------------
--
-- GREATEST function with four parameters

420 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle4
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 >= P2
THEN CASE
WHEN P1 >= P3
THEN CASE
WHEN P1 >= P4 THEN P1
ELSE P4
END
ELSE CASE
WHEN P3 >= P4 THEN P3
ELSE P4
END
END
ELSE CASE
WHEN P2 >= P3
THEN CASE
WHEN P2 >= P4 THEN P2
ELSE P4
END
ELSE CASE
WHEN P3 >= P4 THEN P3
ELSE P4
END
END
END
;
---------------------------------------------------
--
-- GREATEST function with five parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254), P5
VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle5
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 >= P2
THEN CASE
WHEN P1 >= P3
THEN CASE
WHEN P1 >= P4
THEN CASE
WHEN P1 >= P5 THEN P1
ELSE P5
END
ELSE CASE
WHEN P4 >= P5 THEN P4
ELSE P5
END
END
ELSE CASE
WHEN P3 >= P4

Appendix B. Sample code for user defined functions 421

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

THEN CASE
WHEN P3 >= P5 THEN P3
ELSE P5
END
ELSE CASE
WHEN P4 >= P5 THEN P4
ELSE P5
END
END
END
ELSE CASE
WHEN P2 >= P3
THEN CASE
WHEN P2 >= P4
THEN CASE
WHEN P2 >= P5 THEN P2
ELSE P5
END
ELSE CASE
WHEN P4 >= P5 THEN P4
ELSE P5
END
END
ELSE CASE
WHEN P3 >= P4
THEN CASE
WHEN P3 >= P5 THEN P3
ELSE P5
END
ELSE CASE
WHEN P4 >= P5 THEN P4
ELSE P5
END
END
END
END
;
---------------------------------------------------
--
-- GREATEST function with six parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle6
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
GREATEST(GREATEST(P1,P2,P3),GREATEST(P4,P5,P6))
;
---------------------------------------------------
--
-- GREATEST function with seven parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle7
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN

422 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

GREATEST(GREATEST(P1,P2,P3,P4),GREATEST(P5,P6,P7))
;
---------------------------------------------------
--
-- GREATEST function with eight parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254), P8 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle8
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
GREATEST(GREATEST(P1,P2,P3,P4),GREATEST(P5,P6,P7,P8))
;
---------------------------------------------------
--
-- GREATEST function with nine parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254), P8 VarChar(254)
, P9 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle9
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
GREATEST(GREATEST(P1,P2,P3,P4,P5),GREATEST(P6,P7,P8,P9))
;
---------------------------------------------------
--
-- GREATEST function with ten parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION GREATEST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254), P8 VarChar(254)
, P9 VarChar(254),P10 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC GREATESTOracle10
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
GREATEST(GREATEST(P1,P2,P3,P4,P5),GREATEST(P6,P7,P8,P9,P10))
;
---------------------------------------------------

Example B-18 shows the results of UDFs GREATEST.

Example: B-18 Result of UDFs mapping GREATEST

SELECT char(greatest('abcdefg','abcfgh'),20) FROM sysibm.sysdummy1;

---------------------------------------------------

1
--------------------
abcfgh

Appendix B. Sample code for user defined functions 423

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(greatest('abcdefg','defgh','abcfgh'),20) FROM sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
defgh

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(greatest('abcdefg','defgh','abcfgh','endof...'),20) FROM sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
endof...

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(greatest('abcdefg','defgh','abcfgh','endof...','add on'),20) FROM sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
endof...

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(greatest('abcdefg','defgh','abcfgh','endof...','add on','extra'),20) FROM
sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
extra

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(greatest('abcdefg','defgh','abcfgh','endof...','add on','extra','a bit of'),20) FROM
sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
extra

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(greatest('abcdefg','defgh','abcfgh','endof...','add on','extra','a bit of','more'),20)
FROM sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
more

1 record(s) selected.

--------------------------------------------------------------------------

424 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

SELECT char(greatest('abcdefg','defgh','abcfgh','endof...','add on','extra','a bit of','more','more

and '),20) FROM sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
more and

1 record(s) selected.

--------------------------------------------------------------------------
SELECT char(greatest('abcdefg','defgh','abcfgh','endof...','add on','extra','a bit of','more','more
and ',' something'),20) FROM sysibm.sysdummy1;
---------------------------------------------------

1
--------------------
more and

1 record(s) selected.

B.5 Sample code for LEAST

Example B-19 is a set of UDF examples emulating the behavior of MySQL
LEAST function. The various UDFs accept input parameters in varchar and from
2 to 10 input parameters.

Example: B-19 User-defined functions to map LEAST

-- DB2 UDF(User-Defined Function) Samples for conversion

--
-- 2001/08/28, 08/29
--
-- Name of UDF: LEAST (P1 VarChar(254), P2 VarChar(254))
--
--
-- Used UDF: None
--
-- Description: Returns least value of list of data.
--
-- Author: TOKUNAGA, Takashi
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle2
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 <= P2 THEN P1
ELSE P2
END
;
---------------------------------------------------
--
-- LEAST function with three parameters
--

Appendix B. Sample code for user defined functions 425

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle3
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 <= P2
THEN CASE
WHEN P1 <= P3 THEN P1
ELSE P3
END
ELSE CASE
WHEN P2 <= P3 THEN P2
ELSE P3
END
END
;
---------------------------------------------------
--
-- LEAST function with four parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle4
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 <= P2
THEN CASE
WHEN P1 <= P3
THEN CASE
WHEN P1 <= P4 THEN P1
ELSE P4
END
ELSE CASE
WHEN P3 <= P4 THEN P3
ELSE P4
END
END
ELSE CASE
WHEN P2 <= P3
THEN CASE
WHEN P2 <= P4 THEN P2
ELSE P4
END
ELSE CASE
WHEN P3 <= P4 THEN P3
ELSE P4
END
END
END
;
---------------------------------------------------
--
-- LEAST function with five parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254))

426 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle5
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
CASE
WHEN P1 <= P2
THEN CASE
WHEN P1 <= P3
THEN CASE
WHEN P1 <= P4
THEN CASE
WHEN P1 <= P5 THEN P1
ELSE P5
END
ELSE CASE
WHEN P4 <= P5 THEN P4
ELSE P5
END
END
ELSE CASE
WHEN P3 <= P4
THEN CASE
WHEN P3 <= P5 THEN P3
ELSE P5
END
ELSE CASE
WHEN P4 <= P5 THEN P4
ELSE P5
END
END
END
ELSE CASE
WHEN P2 <= P3
THEN CASE
WHEN P2 <= P4
THEN CASE
WHEN P2 <= P5 THEN P2
ELSE P5
END
ELSE CASE
WHEN P4 <= P5 THEN P4
ELSE P5
END
END
ELSE CASE
WHEN P3 <= P4
THEN CASE
WHEN P3 <= P5 THEN P3
ELSE P5
END
ELSE CASE
WHEN P4 <= P5 THEN P4
ELSE P5
END
END
END
END
;
---------------------------------------------------
--
-- LEAST function with six parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)

Appendix B. Sample code for user defined functions 427

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

, P5 VarChar(254), P6 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle6
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LEAST(LEAST(P1,P2,P3),LEAST(P4,P5,P6))
;
---------------------------------------------------
--
-- LEAST function with seven parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle7
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LEAST(LEAST(P1,P2,P3,P4),LEAST(P5,P6,P7))
;
---------------------------------------------------
--
-- LEAST function with eight parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254), P8 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle8
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LEAST(LEAST(P1,P2,P3,P4),LEAST(P5,P6,P7,P8))
;
---------------------------------------------------
--
-- LEAST function with nine parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254), P8 VarChar(254)
, P9 VarChar(254))
RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle9
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LEAST(LEAST(P1,P2,P3,P4,P5),LEAST(P6,P7,P8,P9))
;
---------------------------------------------------
--
-- LEAST function with ten parameters
--
--------------------------------------------------------------------------
CREATE FUNCTION LEAST (P1 VarChar(254), P2 VarChar(254), P3 VarChar(254), P4 VarChar(254)
, P5 VarChar(254), P6 VarChar(254), P7 VarChar(254), P8 VarChar(254)
, P9 VarChar(254),P10 VarChar(254))

428 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

RETURNS VarChar(254)
LANGUAGE SQL
SPECIFIC LEASTOracle10
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN
LEAST(LEAST(P1,P2,P3,P4,P5),LEAST(P6,P7,P8,P9,P10))
;

Example B-20 shows the results of UDF LEAST.

Example: B-20 Results of UDFs mapping LEAST

SELECT least('HARRY','HARRIOT','HAROLD') FROM sysibm.sysdummy1;

---------------------------------------------------

1
---------------------------------------------------------------------------------------
HAROLD

1 record(s) selected

B.6 Sample code for BIT_COUNT

Example B-21 is a UDF example emulating the behavior of MySQL’s
BIT_COUNT function. It returns the number of set bits in the parameter (the
number of 1 in the binary value of the parameter) assuming that the parameter is
a 32-bit INTEGER.

Example: B-21 User-defined function to map BIT_COUNT

CREATE FUNCTION BIT_CNT (N1 Integer)

RETURNS Integer
SPECIFIC BITCNTMySQL
LANGUAGE SQL
CONTAINS SQL
NO EXTERNAL ACTION
DETERMINISTIC
RETURN
WITH
Repeat (S, M1, Ans) AS
(Values (0, N1, 0)
Union All
Select S+1, M1/2, Ans+MOD(M1,2)
From Repeat
Where M1 <> 0
AND S < 32
)
SELECT case when ANS > 0 then ANS else 32 + ANS end
FROM Repeat
WHERE S = (SELECT MAX(S)
FROM Repeat)
;

Appendix B. Sample code for user defined functions 429

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

Example B-22 shows the sample output.

Example: B-22 Sample output of BIT_CNT

db2> values bit_cnt(64)

1
-----------
1

1 record(s) selected.

db2> values bit_cnt(63)

1
-----------
6

1 record(s) selected.

db2> values bit_cnt(-7)

1
-----------
29

1 record(s) selected.

B.7 Sample code for SUBSTRING_INDEX

Example B-23 is a UDF example emulating the behavior of MySQL’s
SUBSTRING_INDEX function. It returns the substring from input string before
counting occurrences of the delimiter.

Example: B-23 User-defined function to map SUBSTRING_INDEX

create function SUBSTRING_INDEX(In varchar(2000),delimit varchar(200), n Int)

returns varchar(2000)
deterministic no external action contains sql
begin atomic
declare out varchar(2000);
declare dem varchar(2000);
declare num int;
declare pos int;
declare temp varchar(2000);
set dem=delimit;
set temp=In;
set num=n;
set pos=1;
if(num<0) then
while(locate(delimit,temp)!=0) do
set temp=substr(temp,locate(delimit,temp)+1);
set num=num+1;
end while;
set num=num+1;
set temp=In;
end if;
while (num>0) do
set pos=pos+locate(delimit,temp)-1;

430 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093ax02.fm

set temp=substr(temp,locate(delimit,temp)+1);
set num=num-1;
end while;
if(n>0) then
return substr(In,1,pos);
else
return substr(In,pos+1);
end if;
end

Appendix B. Sample code for user defined functions 431

7093ax02.fm Draft Document for Review October 16, 2009 10:21 am

432 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093bibl.fm

Related publications

The publications listed in this section are considered particularly suitable for a
more detailed discussion of the topics covered in this book.

IBM Redbooks
For information about ordering these publications, see “How to get Redbooks” on
page 437. Note that some of the documents referenced here may be available in
softcopy only.
򐂰 Data Servers Developing PHP Applications for IBM Data Servers,
SG24-7218
򐂰 MySQL to DB2 UDB Conversion Guide, SG24-7093
򐂰 Up and Running with DB2 on Linux, SG24-6899
򐂰 Oracle to DB2 Conversion Guide for Linux, UNIX, and Windows, SG24-7048

Other publications
These publications are also relevant as further information sources:
򐂰 IEEE Standard for Software Test Documentation (829-1998)
ISBN 0-7381-1444-8
򐂰 Understanding DB2, Learning Visually with Examples, Second Edition,
ISBN-13:978-0-13-158018-3
򐂰 Installing IBM Data Server Clients, GC27-2454-00
򐂰 Installing DB2 Servers, GC27-2455-00
򐂰 Getting Started with DB2 Installation and Administration on Linux and
Windows, GI11-9411-00
򐂰 Database Administration Concepts and Configuration Reference,
SC27-2442-00
򐂰 Database Monitoring Guide and Reference, SC27-2458-00
򐂰 Database Security Guide, SC27-2443-00
򐂰 Partitioning and Clustering Guide, SC27-2453-00

© Copyright IBM Corp. 2009. All rights reserved. 433

7093bibl.fm Draft Document for Review October 16, 2009 10:21 am

򐂰 Troubleshooting and Tuning Database Performance, SC27-2461-00

򐂰 pureXML Guide, SC27-2465-00
򐂰 Data Movement Utilities Guide and Reference, SC27-2440-00
򐂰 Data Recovery and High Availability Guide and Reference, SC27-2441-00
򐂰 Workload Manager Guide and Reference, SC27-2464-00
򐂰 Getting Started with Database Application Development, GI11-9410-00
򐂰 Developing ADO.NET and OLE DB Applications, SC27-2444-00
򐂰 Developing Embedded SQL Applications, SC27-2445-00
򐂰 Developing Java Applications, SC27-2446-00
򐂰 Developing Perl, PHP, Python, and Ruby on Rails Applications,
SC27-2447-00
򐂰 SQL Procedural Languages: Application Enablement and Support,
SC27-2470-00
򐂰 Administrative API Reference, SC27-2435-00
򐂰 Administrative SQL Routines and Views, SC27-2436-00
򐂰 Command Reference, SC27-2439-00
򐂰 Message Reference Vol.1, SC27-2450-00
򐂰 Message Reference Vol.2, SC27-2451-00
򐂰 SQL Reference Vol.1, SC27-2456-00
򐂰 SQL Reference Vol.2, SC27-2457-00
򐂰 XQuery Reference, SC27-2466-00
򐂰 Introduction to Replication and Event Publishing, GC19-1028-02
򐂰 SQL Replication Guide and Reference, SC19-1030-02
򐂰 MySQL 5.0 Certification Study Guide, ISBN-0-672-32812-7

Online resources
These Web sites are also relevant as further information sources:

DB2
򐂰 Database Management
http://www.ibm.com/software/data/management/
򐂰 DB2

434 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093bibl.fm

http://www.ibm.com/software/data/db2/
򐂰 DB2 Express-C
http://www.ibm.com/software/data/db2/express/
http://www.ibm.com/developerworks/forums/forum.jspa?forumID=805
򐂰 IBM DB2 Database for Linux, UNIX, and Windows Information Center
http://publib.boulder.ibm.com/infocenter/db2luw/v9r7/index.jsp
򐂰 DB2 Application Development
http://www.ibm.com/software/data/db2/ad/
򐂰 DB2 Bootcamp Training
http://www.ibm.com/developerworks/data/bootcamps
򐂰 DB2 Linux Validation
http://www.ibm.com/software/data/db2/linux/validate/
򐂰 DB2 9.7 Manuals
http://www1.ibm.com/support/docview.wss?rs=71&uid=swg27015148
򐂰 DB2 9.7 Features and benefits
http://www-01.ibm.com/software/data/db2/9/features.html
򐂰 DB2 Migration Now
http://www.ibm.com/db2/migration
򐂰 IBM Data Movement Tool
http://www.ibm.com/developerworks/data/library/techarticle/dm-0906datamovem
ent/
򐂰 DB2 Technical Support
http://www.ibm.com/software/data/db2/support/db2_9/
򐂰 Integrated Data Management
http://www.ibm.com/software/data/optim/
򐂰 IBM Developerworks
http://www.ibm.com/developerworks/
򐂰 IBM PartnerWorld
http://www.ibm.com/partnerworld
򐂰 Software Migration Project Office
http://www.ibm.com/software/solutions/softwaremigration/
򐂰 Leveraging MySQL skills to learn DB2 Express: DB2 versus MySQL
administration and basic tasks
http://www.ibm.com/developerworks/db2/library/techarticle/dm-0602tham2/

Related publications 435

7093bibl.fm Draft Document for Review October 16, 2009 10:21 am

򐂰 Leverage MySQL skills to learn DB2 Express: DB2 versus MySQL backup
and recovery, viewed June 28, 2008,
http://www.ibm.com/developerworks/db2/library/techarticle/dm-0606tham/
򐂰 Leverage MySQL skills to learn DB2 Express, Part 3: DB2 versus MySQL
graphical user interface, viewed June 28, 2008,
http://www.ibm.com/developerworks/db2/library/techarticle/dm-0608tham/
򐂰 Leverage MySQL skills to learn DB2 Express, Part 4: DB2 versus MySQL
data movement, viewed June 28, 2008,
http://www-128.ibm.com/developerworks/db2/library/techarticle/dm-0610tham/
򐂰 Convert from MySQL or PostSQL to DB2 Express-C, viewed June 28, 2008,
http://www.ibm.com/developerworks/db2/library/techarticle/dm-0606khatri/
򐂰 DB2 Basics: Fun with Dates and Times, viewed August 25th
http://www.ibm.com/developerworks/data/library/techarticle/0211yip/0211yip3
.html

MySQL
򐂰 MySQL home page
http://www.mysql.com/
򐂰 MySQL 5.1 Reference Manual
http://dev.mysql.com/doc/refman/5.1/en/index.html
򐂰 PHP MyAdmin
http://www.phpmyadmin.net/home_page/index.php

Others
򐂰 VMware
http://www.vmware.com/
򐂰 SUSE Linux Enterprise
http://www.novell.com/linux/
򐂰 PHP
http://www.php.net/
򐂰 PHP PECL extension
http://pecl.php.net/
򐂰 PHP Manual - Database extensions
http://ca2.php.net/manual/en/refs.database.php
򐂰 APACHE

436 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:21 am 7093bibl.fm

http://www.apache.org/
򐂰 Perl
http://www.perl.org/
򐂰 Comprehensive Perl Archive Network
http://www.cpan.org
򐂰 Ruby
http://www.ruby-lang.org/en/
򐂰 IBM and Ruby
http://rubyforge.org/projects/rubyibm
򐂰 MySQL and Ruby
http://rubyforge.org/projects/mysql-ruby/
http://tmtm.org/en/ruby/mysql/README_en.html
http://www.tmtm.org/en/mysql/ruby/
򐂰 Java.sql Package Documentation
http://java.sun.com/j2se/1.4.2/docs/api/java/sql/package-summary.html
򐂰 UnixODBC
http://www.unixodbc.org/

How to get Redbooks

You can search for, view, or download Redbooks, Redpapers, Technotes, draft
publications and Additional materials, as well as order hardcopy Redbooks
publications, at this Web site:
ibm.com/redbooks

Help from IBM

IBM Support and downloads
ibm.com/support

IBM Global Services

ibm.com/services

Related publications 437

7093bibl.fm Draft Document for Review October 16, 2009 10:21 am

438 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:25 am 7093IX.fm

Index
application flow 76
Symbols application function 56
.FRM file 124
application interface 37, 56
.MRG file 124
application porting 57
.MYD file 124
application profile 56
.MYI file 124
application server 26
application user account 177
Numerics archive storage engine 46
2-tier 25 ASNCLP command 298
32-bit 4 assignment 215
64-bit 4 async i/o 14
audit statement 183
authentication 180
A autocommit 275
access control 177
automated backup 5
access package 29
automatic storage 127
access path 17
autonomic computing daemon 10
access plan 28, 353
autonomic management feature 5
access right 167
accessctrl authority 183, 187
accesslist table 269 B
account detail 79 background process 10
ActiveX Data Object 32 backup compression 5
address space 9 bash shell 146
administrative interface 37 batch application 25
administrative views 347 big integer 118
ADO 32 binary data 175
ADO.NET 32 binary large object 175
aggregate function 223 bind command 28
all privileges privilege 188 bind commands 277
alter privilege 188 bindadd privilege 187
alterin privilege 188 binding 27
AMD workstation 4 blackhole storage engine 47
ANSI mode 212 BLOB 175
ANSI92 standard 219 block device 14
applaccess table 269 buffer pool 14, 164
applet 30 buffers 39
application architecture 56, 62 built-in data 116
application assessment 62 built-in function 160
application client 26 business intelligence 6, 57
application data 68
application developer 2
application development 92
C
C client library 38
application environment 56
C/C++ 62

© Copyright IBM Corp. 2009. All rights reserved. 439

7093IX.fm Draft Document for Review October 16, 2009 10:25 am

cache 39 control center 90

cache size 127 control privilege 186, 188
Cartesian product 210 conversion task 143
catalog 18 Copy index data 173
catalog command 24 core file 335
catalog table space 128 crash recovery 38, 129, 292
catalog view 13 create procedure syntax 161
certification 60 create trigger statement 160
character data type 218 create_external_routine privilege 187
check constraint 120 create_not_fenced_routine privilege 187
check pending state 326 createin privilege 188
client code 9 createtab privilege 187
client connection 91 cross join 210
client layer 37 CSV storage engine 46
client library 249 cursor stability 274
client support 91 custom installation type 90
client-server application 7 CVS 23
comma separated value 46 CVS storage engine 130
command line interface 66
command line mode 149
commercial license 36
D
dascrt 103
commit 129
dasupdt 101
common client 3
data compression 2
communication 88
data definition language 67
communication encryption 8
data dictionary 41
communication protocol 91
data extraction 144
compact installation type 90
data file 41
compile time 27
data manipulation language 71
compressed format 46
data movement utility 281
compressed row 16
data page 14
compressed table 46
data porting 56
compression rate 58
data recovery 281
concurrency 71
data recovery module 57
concurrent insert 129
data server 2, 91
concurrent versions system 23
data type mapping 115
condition handling 260
data types 175
configuration file 11, 108
data warehouse 58
configuration parameter 127
dataaccess authority 184, 187
configure command 106
database 11
configure script 106
database activity 91
connect privilege 187
database administration 38, 103
connection pool 38
database administrator 2, 59
connection protocol 37
database architecture 62
connection statement 225
database authority 183
connection string 234
database backup 72
console 144
database configuration 281
console mode 152
database configuration parameters
container 14
database_memory 307
context switch 9

440 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:25 am 7093IX.fm

locklist 307 db2imigr 19

logarchmeth1 289 db2import 302
logarchmeth2 289 db2isetup 20
maxlocks 307 db2iupd 101
pckcachesz 307 db2iupdt 19
sheapthres_shr 307 db2level 20
sortheap 308 db2load 305
trackmod 289 db2look 20, 325
database design 60 db2ls 334
database discovery 25 db2pd 20, 334
database driver 224, 234 db2rc 298
database element 144 db2set 283
database feature 2 db2setup 20, 101
database interface 62 db2sqljcustomize 279
database managed space 13 db2start 19
database management system 11, 57 db2stop 19
database management utility 39 DB2 dump file 335
database manager 28 DB2 privilege level
database manager configuration parameter 127 database l 187
database modeling tool 147 row or column 188
database node 125 table space 188
database object 13, 18, 123 DB2 system controller 10
database partition 12, 125 DB2 watch dog 10
database partition group 12 db2comm 24
database recovery 287 db2diag tool 333
database replication 281, 297 db2diag.log 335
database server 38 db2fmp 10
database structure 144 db2greg tool 333
database system monitor 181 db2isetup command 102
database user account 177 db2jcc4.jar 23, 30
date and time data type db2level command 333
date 118 db2move utility 303
datetime 119 db2setup.err 102
time 119 db2setup.log 102
timestamp 119 DB2Sqlca 264
year 119 db2support 334
DB2 commands db2sysc 10
b2admin 20 db2systm file 284
dasauto 20 db2wdog 10
dascrt 20 DBADM authority 182, 187
dasdrop 20 dbheap 364
dasmigr 20 DBI 33
db2acd 10 DBI interface 224
db2ca 104, 285 DBMS 57
db2cc 20 DDL 67, 122
db2dart 20, 333 deadlocks 274
db2icrt 19, 103 decompress table 16
db2idrop 19 default password 77
db2ilist 19, 283 delete privilege 188

Index 441
7093IX.fm Draft Document for Review October 16, 2009 10:25 am

delimited ASCII format file 299 foreign key constraint 47

design advisor 5 foriegn key 141
desktop system 7 formatted page 173
destination server 86 frm extension 43
device name 14 full table scan 17
diagnostic level 283
direct I/O 14
directory name 14
G
global level profile registry 282
discover_db parameter 25
global levels 282
discover_inst parameter 25
global variable Privilege 183
disk space 90
go to label 265
distinctrow keyword 209
grant command 179
distributed platform 2
grant table buffer 39
DML 71
granting privilege 70
double precision 118
graphical interface 38
driver code 3
graphical tool 90
dropin privilege 188
dynamic query 187
dynamic SQL 187 H
dynamic sql statement 28 hashed index 133
dynamic table 46 health and fault monitor 5
dynamic warehouse 7 heap size 127
heap storage engine 130
heap table 46
E High availability 281
e-business 57
high availability 4
EDU 10
host information 24
education opportunity 60
host language variable 27
EJB 241
host name 39
embedded analytical feature 7
embedded SQL statement 27
embedded system 3 I
employee inventory 76 I/O bound workload 16
encrypted password 168 ibm_db2 228
engine dispatchable unit 10 implicit casting 215
engine infrastructure 10 implicit privilege 186
Enterprise JavaBean 241 implicit_schema privilege 187
execute privilege 184 import command 168
explain authority 183, 187 import mode 303
export mode 303 IMS 2
incremental backup 289
independent software vendor 5
F index 136
failover 57
index blocks 39
federated storage engine 47, 130
Index privilege 183
file name 14
index privilege 188
file system 89, 123
indexe 13
fixed term license 5
InnoDB engine 162
floating-point number 118
InnoDB table 124
foreign key 48, 162

442 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:25 am 7093IX.fm

insert privilege 188 load authority 184

installation method 91 load mode 303
installation option 101 load privilege 187
installation program 90 local connection 23
installation status 101 local tools catalog 100
installation type 90 localhost 149
installFixPak command 101 lock escalation 274
instance 11 lock time-out 274
instance level profile registry 282 lock wait 274
instance node level profile registry 282 locking 71
instance owner 193 locklist 164
instance owner information 98 log file 101, 199
integrated exchange format 299 log file path 127
integrity information 18 log files location 283
interactive deployment mode 153 log information 41
interprocess communication 23 logical model 68
inventory 77 loopback 23
inventory location 82
inventory type 82
IP address 229
M
main process 10
IPC 23
mainframe 2
isolation level 71, 273, 277
maintenance window 164
isolation option 277
make command 106
manageability 59
J management tool 59
Java 62 materialized query 133
Java code 31 memory 90
Java enabled browser 30 memory cache 39
JavaServer Page 241 memory model 4
join order 208 memory storage engine 46
JSP 241 memory table 133
merge storage 46
merge storage engine 130
K merge table 124
key buffer 39
method privileges 188
migration assessment 56
L migration project 56
label based access control 2, 179 migration task 56
language flag 102 mincommit 364
large object 13 mobile device 8
LBAC 179 module 184
LBAC rule exemption privilege 188 module privilege 183
licensing model 36 mounting option 123
lightweight deployment solution 23 MQT 134
lightweight security audit mechanism 58 multi-byte character set 33
Linux distribution 88 multiple database 140
list command 141 multi-user version 5
ln command 107 mv command 109

Index 443
7093IX.fm Draft Document for Review October 16, 2009 10:25 am

MYD extension 43 mysqlaccess 50

MYI extension 43 mysqladmin 50
MyISAM table 124 mysqlbinlog 50
myisamchk 50 mysqlcheck 50
myisampack 50 mysqldump 40, 50, 144, 168
mysql 50 mysqldump options
MySQL Administrator 50 --help 169
MySQL index --no-create-info 169–171
fulltext 136 --no-data 169
non-unique 136 --password 169
primary key 136 --tab 170
spatial 136 --tab= 169
unique 136 --user 169
MySQL Query Browser 50 mysqlhotcopy 40, 50, 172
MySQL server program mysqlhotcopy script 168
comp_err 50 mysqlimport 50
innochecksum 50 mysqlshow 50
make_binary_distribution 50
make_win_bin_dist 50
msql2mysql 50
N
named pipe 39
my_print_defaults 50
national language support 71
myisamlog 50
natural join 210
mysql.server 49
nickname privilege 183
mysql_config 50
node 24, 125
mysql_fix_privilege_tables 50
non-recoverable database 288
mysql_install_db 50
non-transaction-safe storage engine 45
mysql_secure_installation 50
null value 219
mysql_tzinfo_to_sql 50
numeric data type 218
mysql_upgrade 50
bigint 118
mysqlbug 50
bit 117
mysqld 49
bool 117
mysqld_multi 49
boolean 117
mysqld_safe 49
decimal 118
mysqld-debug 49
double 118
mysqld-max 49
fixed 118
mysqld-nt 49
float 118
mysqlmanager 49
int 118
perror 50
integer 118
replace 50
numeric 118
mysql.columns_priv table 179
real 118
mysql.procs_priv table 179
smallint 117
mysql.tables_priv table 179
tinyint 117
mysql.user table 177, 179
numeric values 217
mysql_convert_table_format 50
mysql_fix_extensions 50
mysql_setpermissions 50 O
mysql_tableinfo 50 object privileges 18
mysql_waitpid 50 object-oriented extension 107
mysql_zap 50 ODBC driver manager 29

444 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:25 am 7093IX.fm

OLAP 7, 57 database 179

OLE DB provider 32 global level 179
OLTP 57 routine 179
on-line analytical processing 57 table 179
online memory tuning 164 procedural interface 107
on-line transaction processing 57 process identifier 338
optimizer 39, 208 profile 180
profile registry 283
profile registry variables 282
P pureXML, autonomics 2
package 65
package cache 164
package cache size 307 Q
package privilege 183, 188 query cache 39
parallel database system 11 query interface 37
parse tree 39 query optimization class 208
Parser 38 query performance 134
partition group 126 query syntax 39
partitioned database environment 125 query users 37
password 77 quiesce_connect privilege 187
payload file deployment 92
PDO 227
pdo_ibm 33
R
range-clustered table 135
performance 9
raw device 14
Perl 62
raw disk partitions 47
permission 186
RDO 32
PHP 62
read stability 274
PHP data object 227
rebuildconf command 110
PHP source 105
recover history file 288
phpinfo() 107
recoverable database 289
physical resource 39
recovery log 125
physical storage device 14
recovery log file 288
physical structure 11
recovery logs file 287
plug-in 23
recovery method 287
policy 311
Redbooks Web site 437
port address 24
Contact us xv
port number 103
reference type 395
Porting preparation 56
REFERENCES privilege 188
precision 117
Referential integrity 48
precompile option 273
referential integrity 162
precompiled package 107
register variable
pre-compiling 27
db2_compatibility_vector 221
PREP 277
registered user 77
pre-parsed tree 11
registry variable 24, 221, 282
primary key 162
relational modeling 2
primary keys 65, 141
remote connection 23, 36
privilege 13, 167, 186
remote data object 32
privilege level
repeatable read 274
column 179
replication service 4

Index 445
7093IX.fm Draft Document for Review October 16, 2009 10:25 am

reserved words 145 slave database 40

resource tuning database technology 59 slave system 40
response file 101 small integer 117
response file installation 92 sort heap 164
restore utility 295 sort heap size 308
result output file 144 sort heap threshold 307
revoke command 179 source code language 62
roles 179 source distribution 41
roll forward recovery 293 source table 203
rollback 129 sources package 109
rollforward utility 296 special group 186
root privilege 180 SQL 2
root table 135 SQL interface 38
root type 395 SQL replication 5
routine privilege 183 SQL statements 183
routines 184 sqladm authority 183, 187
row compression 131, 384 SQLCODE 260
row trigger 137 SQLj 241
rpm package 41 SQLj applet 31
runstats command 199 SQLj translator 31
runtime support 23 SQLSTATE 260
standard interface 25
statement trigger 137
S static sql statement 27
schema 43, 128
static tables 46
Schema level 187
statistical data 41
Schema privilege 183
status 144
secadm authority 183, 187
STMM 5, 59, 164
security 11, 13
storage device 126
security label privilege 188
storage engine 39–40, 129
security model 180
storage engines 44
security system 177
storage optimization 4
select privilege 188
stored procedure 22, 84
self tuning memory manager 5
straight_join keyword 208, 210
self-tuning memory manager 59, 164
string and character data type
sequence privilege 183, 188
binary 119
server code 9
blob 120
server privilege 183
char 119
service request 77
enum 120
service ticket 76, 81
longblob 120
service type 82
longtext 120
servlets 241
mediumblob 120
setup wizard 89–90
mediumtext 120
severity level 84
set 120
shared memory networking protocol 39
tinyblob 119
shared-nothing architecture 47
tinytext 120
single byte integer 117
varbinary 119
single precision 118
varchar 119
single-tier 25
string function 213

446 MySQL to DB2 Conversion Guide

Draft Document for Review October 16, 2009 10:25 am 7093IX.fm

strong typing 215 transaction-safe storage engines 45

structured query language 123 TRG extension 43
structured type 395 trigger 65, 137
subquery 211 TRN extension 43
subtable 135 truncate statement 220
summary table 134 two-tier architecture 63
supertable 135 type 4 driver 30
symbolic link 123 typed tables 135
synchronous event 14 typical installation type 90
synonym 117, 209
sysadm authority 181
syscat catalog view 18
U
uncommitted read 274
sysctrl authority 181
underscore 178
sysmaint authority 181
Unified ODBC 235
sysmon authority 181
unique key 141, 162
sysstat catalog view 18
unit of work 277
system catalog 11
unload script 172
system catalog table 13, 128
update privilege 188
system directory 11
updateable view 174
system managed space 13
use privilege 188
system monitor 181
user account 77, 98
system planning 60
user account information 39, 177
system requirement 89
user account management 167, 177
User data 68
T user data 167
table column 11 user defined function 10
table descriptor 39 user defined table 128
table hierarchy 135 user privilege 40
table partitioning 2 user table space 128
table space change history file 288 user-defined data type 116
table space privilege 183 user-defined partitioning 124
table spaces 12
table-level encryption 8
tar file 105
V
validation program 88
target system 56, 63
value compression 384
TCP/IP communication 100
vector I/O 14
temporary table 128
version recovery 292
temporary table space 128
view 13
text 144
threaded engine 2
threaded model 9 W
throttled utility 311 warehouse architect 59
throttling system 311 Web server 26
timestamp 217 Web service 22
trace file 101 Web-browser 26
transaction isolation 272 whenever statement 264
transaction processing 57 wildcard 178
transaction safe 129 wlmadm authority 183, 187

Index 447
7093IX.fm Draft Document for Review October 16, 2009 10:25 am

work sheet format 299

workload 164
workload management 2
workload manager 281
workload privilege 183

X
XML support 71
XSR object privilege 184

448 MySQL to DB2 Conversion Guide

To determine the spine width of a book, you divide the paper PPI into the number of pages in the book. An example is a 250 page book using Plainfield opaque 50#
smooth which has a PPI of 526. Divided 250 by 526 which equals a spine width of .4752". In this case, you would use the .5” spine. Now select the Spine width for
the book and hide the others: Special>Conditional Text>Show/Hide>SpineSize(-->Hide:)>Set . Move the changed Conditional text settings to all files in your
book by opening the book file with the spine.fm still open and File>Import>Formats the Conditional Text Settings (ONLY!) to the book files.
Draft Document for Review October 16, 2009 10:21 am 7093spine.fm 449
MySQL to DB2 Conversion Guide
(1.0” spine)
0.875”<->1.498”
460 <-> 788 pages
To determine the spine width of a book, you divide the paper PPI into the number of pages in the book. An example is a 250 page book using Plainfield opaque 50#
smooth which has a PPI of 526. Divided 250 by 526 which equals a spine width of .4752". In this case, you would use the .5” spine. Now select the Spine width for
the book and hide the others: Special>Conditional Text>Show/Hide>SpineSize(-->Hide:)>Set . Move the changed Conditional text settings to all files in your
book by opening the book file with the spine.fm still open and File>Import>Formats the Conditional Text Settings (ONLY!) to the book files.
Draft Document for Review October 16, 2009 10:21 am 7093spine.fm 450
 Back cover ®
Draft Document for Review October 16, 2009 2:07 pm

MySQL to DB2 Conversion

Guide
®

Complete guide to Switching database vendors is often considered an

convert MySQL exhausting challenge for database administrator and INTERNATIONAL
database and developer. Complexity, total cost, and the risk of downtime TECHNICAL
application to DB2 are often the reasons that restrain IT decision makers to start SUPPORT
the migration project. The primary goal of this book is to ORGANIZATION
show that, with the proper planning and guidance,
Application
converting from MySQL to IBM DB2 is not only feasible but
enrichment through
straight forward.
advanced DB2 BUILDING TECHNICAL
features If you picked up this book, you are most likely considering INFORMATION BASED ON
converting to DB2 and are probably aware of some of the PRACTICAL EXPERIENCE
advantages of to converting to DB2 data server. In this IBM
Application
Redbooks publication we discuss in detail how you can take
conversion with IBM Redbooks are developed by
advantage of this industry leading database server.
detailed examples the IBM International Technical
This book is an informative guide that describes how to Support Organization. Experts
convert the database system from MySQL 5.1 to DB2 V9.7 on from IBM, Customers and
Linux, and the steps involved in enabling the applications to Partners from around the world
create timely technical
use DB2 instead of MySQL. information based on realistic
This guide also presents the best practices in conversion scenarios. Specific
recommendations are provided
strategy and planning, conversion tools, porting steps, and to help you implement IT
practical conversion examples. It is intended for technical solutions more effectively in
staff involved in a MySQL to DB2 conversion project. your environment.

For more information:

ibm.com/redbooks

SG24-7093-01 ISBN