The Data Access Handbook

John Goodson: As the executive leader of DataDirect Technologies, John is responsible for daily operations, business development, product direction, and long-term corporate strategy.   John was a principal engineer at Data General for seven years, working on their relational database product, DG/SQL. Since joining DataDirect Technologies in 1992, he has held positions of increasing responsibility in research and development, technical support, and marketing. John is a well-known and respected industry luminary and data connectivity expert. For more than 15 years, he has worked closely with Sun Microsystems and Microsoft on the development and evolution of database connectivity standards including J2EE, JDBC, .NET, ODBC, and ADO. John has been involved with the ANSI NCITS H2 Committee, which is responsible for building the SQL standard, and the X/Open (Open Group) SQL Access Group, which is responsible for building call-level interfaces into relational databases. He is actively involved in Java standards committees, including the JDBC Expert Group. In addition, John has published numerous articles and spoken publicly on topics related to data management. John is also a patent holder in the area of distributed transactions for Microsoft SQL Server Java middleware.   John holds a Bachelor of Science in computer science from Virginia Polytechnic Institute and State University in Blacksburg,Virginia.   Rob Steward: As vice president of research and development for DataDirect Technologies, Rob is responsible for the development, strategy, and oversight of the company’s data connectivity products, including the Shadow mainframe integration suite client software and the industry-leading DataDirect Connect family of database drivers and data providers: Connect for ODBC, Connect for JDBC, and Connect for ADO.NET. Additional product development responsibilities include DataDirect Sequelink and DataDirect XQuery, as well as the management of DataDirect’s Custom Engineering Development group.   Rob has spent more than 15 years developing database access middleware, including .NET data providers, ODBC drivers, JDBC drivers, and OLE DB data providers. He has held a number of management and technical positions at DataDirect Technologies and serves on various standards committees. Earlier in his career, he worked as lead software engineer at Marconi Commerce Systems Inc.   Rob holds a Bachelor of Science in computer science from North Carolina State University, Raleigh.  

Preface

CHAPTER 1 Performance Isn’t What It Used To Be

Where Are We Today?

    The Network

    The Database Driver

    The Environment

    Your Database Application

Our Goal for This Book

CHAPTER 2 Designing for Performance: What’s Your Strategy?

Your Applications

    Database Connections

    Transaction Management

    SQL Statements

    Data Retrieval

    Extended Security

Static SQL Versus Dynamic SQL

The Network

The Database Driver

Know Your Database System

Using Object-Relational Mapping Tools

Summary

CHAPTER 3 Database Middleware: Why It’s Important

What Is Database Middleware?

How Database Middleware Affects Application Performance

Database Drivers

    What Does a Database Driver Do?

    Database Driver Architecture

    Runtime Performance Tuning Options

    Configuring Database Drivers/Data Providers

Summary

CHAPTER 4 The Environment: Tuning for Performance

Runtime Environment (Java and .NET)

    JVM

    .NET CLR

Operating System

Network

    Database Protocol Packets

    Network Packets

    Configuring Packet Size

    Analyzing the Network Path

    Reducing Network Hops and Contention

    Avoiding Network Packet Fragmentation

    Increasing Network Bandwidth

Hardware

    Memory

    Disk

    CPU (Processor)

    Network Adapter

    Virtualization

Summary

CHAPTER 5 ODBC Applications: Writing Good Code

Managing Connections

    Connecting Efficiently

    Using Connection Pooling

    Establishing Connections One at a Time

    Using One Connection for Multiple Statements

    Obtaining Database and Driver Information Efficiently

Managing Transactions

    Managing Commits in Transactions

    Choosing the Right Transaction Model

Executing SQL Statements

    Using Stored Procedures

    Using Statements Versus Prepared Statements

    Using Arrays of Parameters

    Using the Cursor Library

Retrieving Data

    Retrieving Long Data

    Limiting the Amount of Data Retrieved

    Using Bound Columns

    Using SQLExtendedFetch Instead of SQLFetch

    Determining the Number of Rows in a Result Set

    Choosing the Right Data Type

Updating Data

    Using SQLSpecialColumns to Optimize Updates and Deletes

Using Catalog Functions

    Minimizing the Use of Catalog Functions

    Avoiding Search Patterns

    Using a Dummy Query to Determine Table Characteristics

Summary

CHAPTER 6 JDBC Applications: Writing Good Code

Managing Connections

    Connecting Efficiently

    Using Connection Pooling

    Establishing Connections One at a Time

    Using One Connection for Multiple Statements

    Disconnecting Efficiently

    Obtaining Database and Driver Information Efficiently

Managing Transactions

    Managing Commits in Transactions

    Choosing the Right Transaction Model

Executing SQL Statements

    Using Stored Procedures

    Using Statements Versus Prepared Statements

    Using Batches Versus Prepared Statements

    Using getXXX Methods to Fetch Data from a Result Set

    Retrieving Auto-Generated Keys

Retrieving Data

    Retrieving Long Data

    Limiting the Amount of Data Retrieved

    Determining the Number of Rows in a Result Set

    Choosing the Right Data Type

    Choosing the Right Cursor

Updating Data

    Using Positioned Updates, Inserts, and Deletes (updateXXX Methods)

    Using getBestRowIdentifier() to Optimize Updates and Deletes

Using Database Metadata Methods

    Minimizing the Use of Database Metadata Methods

    Avoiding Search Patterns

    Using a Dummy Query to Determine Table Characteristics

Summary

CHAPTER 7 .NET Applications: Writing Good Code

Managing Connections

    Connecting Efficiently

    Using Connection Pooling

    Establishing Connections One at a Time

    Disconnecting Efficiently

    Obtaining Database and Data Provider Information Efficiently

Managing Transactions

    Managing Commits in Transactions

    Choosing the Right Transaction Model

Executing SQL Statements

    Executing SQL Statements that Retrieve Little or No Data

    Using the Command.Prepare Method

    Using Arrays of Parameters/Batches Versus Prepared Statements

    Using Bulk Load

    Using Pure Managed Providers

Selecting .NET Objects and Methods

    Avoiding the CommandBuilder Object

    Choosing Between a DataReader and DataSet Object

    Using GetXXX Methods to Fetch Data from a DataReader

Retrieving Data

    Retrieving Long Data

    Limiting the Amount of Data Retrieved

    Choosing the Right Data Type

Updating Data

Summary

CHAPTER 8 Connection Pooling and Statement Pooling

Connection Pool Model for JDBC

    Configuring Connection Pools

    Guidelines

Connection Pool Model for ODBC

    Connection Pooling as Defined in the ODBC Specification

    Configuring Connection Pools

    Guidelines

Connection Pool Model for ADO.NET

    Configuring Connection Pools

    Guidelines

Using Reauthentication with Connection Pooling

    Configuring Connection Pooling with Reauthentication in a JDBC Environment

Using Statement Pooling

    Using Statement Pooling with Connection Pooling

    Guidelines

Summary: The Big Picture

CHAPTER 9 Developing Good Benchmarks

Developing the Benchmark

    Define Benchmark Goals

    Reproduce the Production Environment

    Isolate the Test Environment

    Reproduce the Workload

    Measure the Right Tasks

    Measure over a Sufficient Duration of Time

    Prepare the Database

    Make Changes One at a Time

    Assess Other Factors

Benchmark Example

Summary

CHAPTER 10 Troubleshooting Performance Issues

Where to Start

Changes in Your Database Application Deployment

The Database Application

The Database Driver

    Runtime Performance Tuning Options

    Architecture

The Environment

    Runtime Environment (Java and .NET)

    Operating System

    Network

    Hardware

Case Studies

    Case Study 1

    Case Study 2

    Case Study 3

    Case Study 4

    Case Study 5

    Case Study 6

    Case Study 7

    Case Study 8

Summary

CHAPTER 11 Data Access in Service-Oriented Architecture (SOA) Environments

What Is Service-Oriented Architecture (SOA)?

Data Access Guidelines for SOA Environments

    Involve Data Experts in Addition to SOA Experts

    Decouple Data Access from Business Logic

    Design and Tune for Performance

    Consider Data Integration

Summary

Glossary

 

 

0137143931    TOC    2/19/2009