Catalogue


Data warehouse : from architecture to implementation /
Barry Devlin.
imprint
Reading, Mass : Addison-Wesley, c1997.
description
xvi, 432 p. : ill. ; 24 cm.
ISBN
0201964252
format(s)
Book
Holdings
More Details
imprint
Reading, Mass : Addison-Wesley, c1997.
isbn
0201964252
catalogue key
822741
 
Includes bibliographical references (p. [411]-415) and index.
A Look Inside
About the Author
Author Affiliation
Barry Devlin--one of the world's leading experts on data warehousing--is also one of the first practitioners in this area. In this book, he distills the insights and experiences gained over 10 years of designing and building data warehouses. Dr. Barry Devlin is a leading authority in Europe on data warehousing. He defined the Data Warehouse Architecture within IBM Europe in 1985 and contributed to its practical implementation over a number of years. This gives him a unique insight into user demands for information, and the development consequences. Barry has a total of 14 years in the IS business, the last 10 of which have been spent with IBM's International Software Development Laboratory in Dublin. He currently works as a consultant in the IBM Consulting Group in Dublin and is a member of the IBM Academy of Technology.
Excerpts
Introduction or Preface
I first conceived of the idea of writing a book on data warehousing during a series of 2-day seminars known as the "Information Warehouse MasterClass" that I and a number of IBM colleagues had developed and presented around the world from 1992 to 1994. From the many companies that attended these seminars, one principal requirement was clear: they needed a common definition or architecture for a data warehouse, detailed enough to drive a consistent implementation within their organizations, yet concise enough to allow the whole company to understand and accept it. It was from the MasterClasses and the needs of these companies that I developed the representation and terminology of the data warehouse architecture used in this book. In 1992, only a few real data warehouse implementations existed, each one hand-crafted and custom-built. Today, the vast majority of companies are planning to build or are actually building a data warehouse. While working with these companies over the years, I and other consultants saw the need to develop methodologies that cover the entire implementation process. This process continues to present major difficulties for many data warehouse implementations. I am convinced that companies today need a generalized and rational implementation approach to this complex process. The methodology described in this book is the result of crafting and proving the implementation approach over the years in real warehouse implementation projects. Clearly, I have distilled the material covered here from interactions with many colleagues within IBM, with other consultants working in the field, and most especially from many hours of work with clients. Without their contributions, this book could not have been written. On the other hand, responsibility for any errors or misinterpretations is, of course, mine. It would be impossible to acknowledge by name everybody who has contributed to this book. To name anybody is to run the risk of omitting some valuable contributions; to those people I apologize in advance. However, I would like to especially thank a number of people either individually or collectively, whose support, knowledge, or time made this book possible: the companies I've worked with over the years, especially those who have agreed to be directly referenced here: Martijn Bossenbroek (ABN AMRO), BØrre Lunde and Solveig Oien Berg (Gjensidige), and Laura Sager (Whirlpool) the colleagues who have contributed to or reviewed the material: John Bair, Peter Cabena, CiarÁn Ennis, Keith Holmes, Edwin Humphreys, Jim McGovern, Paul Murphy, Barry O'Brien, Pat O'Sullivan, Phil Teale, Michael Storey, and Feargal Supple the external reviewers who have added significantly to the book: David Christian, John Kneiling, Eric Rawlins, Richard Rist and Terry Moriarty John Holland and DÓnal O'Shea, who introduced the book to Addison Wesley Longman the team at Addison Wesley Longman: Lynne Doran Cote, Katherine Harutunian, Patty Mahtani and especially my editor Susan Middleton Last, but not least, a special word of thanks to my family, Lil, Katherine, Alan and Emma, who have become convinced over the past year that I can exist only in symbiosis with my PC. 1 Barry Devlin August, 1996 Dublin 1 The text of this book was produced entirely in Microsoft Word and the graphics were developed using Lotus Freelance Graphics. 0201964252P04062001
Introduction or Preface
I first conceived of the idea of writing a book on data warehousing during a series of 2-day seminars known as the "Information Warehouse MasterClass" that I and a number of IBM colleagues had developed and presented around the world from 1992 to 1994. From the many companies that attended these seminars, one principal requirement was clear: they needed a common definition or architecture for a data warehouse, detailed enough to drive a consistent implementation within their organizations, yet concise enough to allow the whole company to understand and accept it. It was from the MasterClasses and the needs of these companies that I developed the representation and terminology of the data warehouse architecture used in this book. In 1992, only a few real data warehouse implementations existed, each one hand-crafted and custom-built. Today, the vast majority of companies are planning to build or are actually building a data warehouse. While working with these companies over the years, I and other consultants saw the need to develop methodologies that cover the entire implementation process. This process continues to present major difficulties for many data warehouse implementations. I am convinced that companies today need a generalized and rational implementation approach to this complex process. The methodology described in this book is the result of crafting and proving the implementation approach over the years in real warehouse implementation projects. Clearly, I have distilled the material covered here from interactions with many colleagues within IBM, with other consultants working in the field, and most especially from many hours of work with clients. Without their contributions, this book could not have been written. On the other hand, responsibility for any errors or misinterpretations is, of course, mine. It would be impossible to acknowledge by name everybody who has contributed to this book. To name anybody is to run the risk of omitting some valuable contributions; to those people I apologize in advance. However, I would like to especially thank a number of people either individually or collectively, whose support, knowledge, or time made this book possible: the companies I've worked with over the years, especially those who have agreed to be directly referenced here: Martijn Bossenbroek (ABN AMRO), Børre Lunde and Solveig Oien Berg (Gjensidige), and Laura Sager (Whirlpool) the colleagues who have contributed to or reviewed the material: John Bair, Peter Cabena, Ciarán Ennis, Keith Holmes, Edwin Humphreys, Jim McGovern, Paul Murphy, Barry O'Brien, Pat O'Sullivan, Phil Teale, Michael Storey, and Feargal Supple the external reviewers who have added significantly to the book: David Christian, John Kneiling, Eric Rawlins, Richard Rist and Terry Moriarty John Holland and Dónal O'Shea, who introduced the book to Addison Wesley Longman the team at Addison Wesley Longman: Lynne Doran Cote, Katherine Harutunian, Patty Mahtani and especially my editor Susan Middleton Last, but not least, a special word of thanks to my family, Lil, Katherine, Alan and Emma, who have become convinced over the past year that I can exist only in symbiosis with my PC.1 Barry Devlin August, 1996 Dublin 1The text of this book was produced entirely in Microsoft Word and the graphics were developed using Lotus Freelance Graphics. 0201964252P04062001
Introduction or Preface
I first conceived of the idea of writing a book on data warehousing during a series of 2-day seminars known as the "Information Warehouse MasterClass" that I and a number of IBM colleagues had developed and presented around the world from 1992 to 1994. From the many companies that attended these seminars, one principal requirement was clear: they needed a common definition or architecture for a data warehouse, detailed enough to drive a consistent implementation within their organizations, yet concise enough to allow the whole company to understand and accept it. It was from the MasterClasses and the needs of these companies that I developed the representation and terminology of the data warehouse architecture used in this book.In 1992, only a few real data warehouse implementations existed, each one hand-crafted and custom-built. Today, the vast majority of companies are planning to build or are actually building a data warehouse. While working with these companies over the years, I and other consultants saw the need to develop methodologies that cover the entire implementation process. This process continues to present major difficulties for many data warehouse implementations. I am convinced that companies today need a generalized and rational implementation approach to this complex process. The methodology described in this book is the result of crafting and proving the implementation approach over the years in real warehouse implementation projects.Clearly, I have distilled the material covered here from interactions with many colleagues within IBM, with other consultants working in the field, and most especially from many hours of work with clients. Without their contributions, this book could not have been written. On the other hand, responsibility for any errors or misinterpretations is, of course, mine. It would be impossible to acknowledge by name everybody who has contributed to this book. To name anybody is to run the risk of omitting some valuable contributions; to those people I apologize in advance. However, I would like to especially thank a number of people either individually or collectively, whose support, knowledge, or time made this book possible: the companies I've worked with over the years, especially those who have agreed to be directly referenced here: Martijn Bossenbroek (ABN AMRO), BØrre Lunde and Solveig Oien Berg (Gjensidige), and Laura Sager (Whirlpool) the colleagues who have contributed to or reviewed the material: John Bair, Peter Cabena, CiarÁn Ennis, Keith Holmes, Edwin Humphreys, Jim McGovern, Paul Murphy, Barry O'Brien, Pat O'Sullivan, Phil Teale, Michael Storey, and Feargal Supple the external reviewers who have added significantly to the book: David Christian, John Kneiling, Eric Rawlins, Richard Rist and Terry Moriarty John Holland and DÓnal O'Shea, who introduced the book to Addison Wesley Longman the team at Addison Wesley Longman: Lynne Doran Cote, Katherine Harutunian, Patty Mahtani and especially my editor Susan MiddletonLast, but not least, a special word of thanks to my family, Lil, Katherine, Alan and Emma, who have become convinced over the past year that I can exist only in symbiosis with my PC. 1Barry Devlin August, 1996 Dublin 1 The text of this book was produced entirely in Microsoft Word and the graphics were developed using Lotus Freelance Graphics. 0201964252P04062001
First Chapter

I first conceived of the idea of writing a book on data warehousing during a series of 2-day seminars known as the "Information Warehouse MasterClass" that I and a number of IBM colleagues had developed and presented around the world from 1992 to 1994. From the many companies that attended these seminars, one principal requirement was clear: they needed a common definition or architecture for a data warehouse, detailed enough to drive a consistent implementation within their organizations, yet concise enough to allow the whole company to understand and accept it. It was from the MasterClasses and the needs of these companies that I developed the representation and terminology of the data warehouse architecture used in this book.

In 1992, only a few real data warehouse implementations existed, each one hand-crafted and custom-built. Today, the vast majority of companies are planning to build or are actually building a data warehouse. While working with these companies over the years, I and other consultants saw the need to develop methodologies that cover the entire implementation process. This process continues to present major difficulties for many data warehouse implementations. I am convinced that companies today need a generalized and rational implementation approach to this complex process. The methodology described in this book is the result of crafting and proving the implementation approach over the years in real warehouse implementation projects.

Clearly, I have distilled the material covered here from interactions with many colleagues within IBM, with other consultants working in the field, and most especially from many hours of work with clients. Without their contributions, this book could not have been written. On the other hand, responsibility for any errors or misinterpretations is, of course, mine. It would be impossible to acknowledge by name everybody who has contributed to this book. To name anybody is to run the risk of omitting some valuable contributions; to those people I apologize in advance. However, I would like to especially thank a number of people either individually or collectively, whose support, knowledge, or time made this book possible:

  • the companies I've worked with over the years, especially those who have agreed to be directly referenced here: Martijn Bossenbroek (ABN AMRO), Brre Lunde and Solveig Oien Berg (Gjensidige), and Laura Sager (Whirlpool)

  • the colleagues who have contributed to or reviewed the material: John Bair, Peter Cabena, Ciarn Ennis, Keith Holmes, Edwin Humphreys, Jim McGovern, Paul Murphy, Barry O'Brien, Pat O'Sullivan, Phil Teale, Michael Storey, and Feargal Supple

  • the external reviewers who have added significantly to the book: David Christian, John Kneiling, Eric Rawlins, Richard Rist and Terry Moriarty

  • John Holland and Dnal O'Shea, who introduced the book to Addison Wesley Longman

  • the team at Addison Wesley Longman: Lynne Doran Cote, Katherine Harutunian, Patty Mahtani and especially my editor Susan Middleton

Last, but not least, a special word of thanks to my family, Lil, Katherine, Alan and Emma, who have become convinced over the past year that I can exist only in symbiosis with my PC. 1

Barry Devlin
August, 1996
Dublin


1 The text of this book was produced entirely in Microsoft Word and the graphics were developed using Lotus Freelance Graphics.


0201964252P04062001

Summaries
Back Cover Copy
Data warehousing is one of the hottest topics in the computing industry today. For business executives, it promises significant competitive advantage for their companies, while information systems managers see it as the way to overcome the traditional roadblocks to providing business information for managers and other end users. With the publication of this book comes the most comprehensive, practical guide to designing, building, and implementing a data warehouse on the market today. Barry Devlin --one of the world's leading experts on data warehousing--is also one of the first practitioners in this area. In this book, he distills the insights and experiences gained over 10 years of designing and building data warehouses. Included are: An explanation of the optimal three-tiered architecture for the data warehouse, with a clear division between data and information A full description of the functions needed to implement such an architecture, including reconciling existing, diverse data and deriving consistent, valuable business information A detailed methodology for building a data warehouse in a way that provides business value and strategic infrastructure at each stage A high-level approach to justifying the effort involved A view of the organizational aspects of building and maintaining a warehouse This book will become the key reference for any team undertaking the construction of a data warehouse. It is aimed primarily at the IS managers, architects, and designers involved in this process, as well as the end users having a key role in the evolving implementation of the data warehouse. 0201964252B04062001
Back Cover Copy
Data warehousing is one of the hottest topics in the computing industry today. For business executives, it promises significant competitive advantage for their companies, while information systems managers see it as the way to overcome the traditional roadblocks to providing business information for managers and other end users. With the publication of this book comes the most comprehensive, practical guide to designing, building, and implementing a data warehouse on the market today.Barry Devlin--one of the world's leading experts on data warehousing--is also one of the first practitioners in this area. In this book, he distills the insights and experiences gained over 10 years of designing and building data warehouses. Included are: An explanation of the optimal three-tiered architecture for the data warehouse, with a clear division between data and information A full description of the functions needed to implement such an architecture, including reconciling existing, diverse data and deriving consistent, valuable business information A detailed methodology for building a data warehouse in a way that provides business value and strategic infrastructure at each stage A high-level approach to justifying the effort involved A view of the organizational aspects of building and maintaining a warehouseThis book will become the key reference for any team undertaking the construction of a data warehouse. It is aimed primarily at the IS managers, architects, and designers involved in this process, as well as the end users having a key role in the evolving implementation of the data warehouse. 0201964252B04062001
Back Cover Copy
Data warehousing is one of the hottest topics in the computing industry today. For business executives, it promises significant competitive advantage for their companies, while information systems managers see it as the way to overcome the traditional roadblocks to providing business information for managers and other end users. With the publication of this book comes the most comprehensive, practical guide to designing, building, and implementing a data warehouse on the market today. Barry Devlin--one of the world's leading experts on data warehousing--is also one of the first practitioners in this area. In this book, he distills the insights and experiences gained over 10 years of designing and building data warehouses. Included are: An explanation of the optimal three-tiered architecture for the data warehouse, with a clear division between data and information A full description of the functions needed to implement such an architecture, including reconciling existing, diverse data and deriving consistent, valuable business information A detailed methodology for building a data warehouse in a way that provides business value and strategic infrastructure at each stage A high-level approach to justifying the effort involved A view of the organizational aspects of building and maintaining a warehouse This book will become the key reference for any team undertaking the construction of a data warehouse. It is aimed primarily at the IS managers, architects, and designers involved in this process, as well as the end users having a key role in the evolving implementation of the data warehouse. 0201964252B04062001
Back Cover Copy
most comprehensive, practical guide to designing, building, and implementing a data warehouse on the market today. Barry Devlin --one of the world's leading experts on data warehousing--is also one of the first practitioners in this area. In this book, he distills the insights and experiences gained over 10 years of designing and building data warehouses. Included are: An explanation of the optimal three-tiered architecture for the data warehouse, with a clear division between data and information A full description of the functions needed to implement such an architecture, including reconciling existing, diverse data and deriving consistent, valuable business information A detailed methodology for building a data warehouse in a way that provides business value and strategic infrastructure at each stage A high-level approach to justifying the effort involved A view of the organizational aspects of building and maintaining a warehouse This book will become the key reference for any team undertaking the construction of a data warehouse. It is aimed primarily at the IS managers, architects, and designers involved in this process, as well as the end users having a key role in the evolving implementation of the data warehouse. 0201964252B04062001
Main Description
Data warehousing is one of the hottest topics in the computing industry. Written by Barry Devlin, one of the world's leading experts on data warehousing, this book gives you the insights and experiences gained over 10 years and offers the most comprehensive, practical guide to designing, building, and implementing a successful data warehouse. Included in this vital information is an explanation of the optimal three-tiered architecture for the data warehouse, with a clear division between data and information. Information systems managers will appreciate the full description of the functions needed to implement such an architecture, including reconciling existing, diverse data and deriving consistent, valuable business information.
Table of Contents
Prefacep. v
Table of contentsp. vii
Table of figures and tablesp. xiii
Introductionp. 1
Why this book?p. 2
Audiencep. 2
Structurep. 3
The evolution of data warehousing
The data warehouse--a brief historyp. 7
Prehistoric times--before the 1980sp. 9
The middle ages--mid- to late-1980sp. 12
The data revolution--the early 1990sp. 15
The era of information-based management--into the 21st centuryp. 18
What is a data warehouse?p. 20
Conclusionsp. 21
Today's development environmentp. 23
Fragmented application developmentp. 24
Operational application developmentp. 24
Application-driven decision supportp. 27
The Info Centerp. 33
Conclusionsp. 35
Principles of data warehousing
Types of data and their usesp. 41
Types of datap. 42
Business datap. 44
Metadatap. 52
Data beyond the scope of the warehousep. 57
Internal and external datap. 59
Conclusionsp. 61
Conceptual data architecturep. 63
Business data architecturesp. 64
The single-layer data architecturep. 64
The two-layer data architecturep. 67
The three-layer data architecturep. 69
A data architecture for metadatap. 77
Conclusionsp. 84
Design techniquesp. 87
Enterprise data modelingp. 88
Representing time in business datap. 97
Historical datap. 104
Data replicationp. 108
Conclusionsp. 123
Introduction to the logical architecturep. 125
Business data in the data warehousep. 126
Business data--other considerationsp. 130
External datap. 134
Metadata in the data warehousep. 137
The data warehouse catalogp. 140
Operational systemsp. 141
Data warehouse functionalityp. 145
Conclusionsp. 148
Creating the data asset
Business data warehouse designp. 151
Modeling the BDW--general designp. 152
Modeling the BDW--a segmented approachp. 155
Modeling the BDW--practical resultsp. 161
The structure of periodic data in the BDWp. 162
Archive and retrievalp. 169
The role of parallel databasesp. 172
Conclusionsp. 174
Populating the business data warehousep. 177
BDW population--initial considerationsp. 178
Capture--an introductionp. 178
From operational data to the BDWp. 180
Six data capture techniquesp. 182
Output data structures from capturep. 194
Apply--an introductionp. 196
Apply during BDW creationp. 197
Apply during BDW maintenancep. 201
Refresh versus update of the BDWp. 204
Transformation--an introductionp. 205
Transformation in BDW populationp. 212
BDW population--the overall processp. 218
Conclusionsp. 219
Unlocking the data asset for end users
Designing business information warehousesp. 223
Types of business information warehousep. 224
Modeling BIWsp. 227
Key influences on BIW designp. 231
BIW implementationp. 234
Historical data in BIWsp. 241
Archive and retrieval in BIWsp. 244
Conclusionsp. 245
Populating business information warehousesp. 247
BIW population--an introductionp. 248
Capture from the BDWp. 249
Apply to the BIWp. 250
Comparing the performance of update and refresh modes of replicationp. 254
Transformationp. 255
BIW population--implementation aspectsp. 259
Conclusionsp. 260
User access to informationp. 261
The business information interfacep. 262
Data accessp. 270
Conclusionsp. 274
Information--data in contextp. 275
The business information guide--an introductionp. 276
Requirements for the BIGp. 277
The naive and sentimental userp. 291
Users of the BIGp. 294
Structure of the BIGp. 296
DWC populationp. 297
Conclusionsp. 298
Implementing the data warehouse
Obstacles to implementationp. 303
The size and scope of the warehousep. 304
Justifying investment in a data warehousep. 305
Organizational issuesp. 305
Placement of the BDW and BIWs in the enterprisep. 306
Ongoing administrationp. 307
Conclusionsp. 307
Planning your implementationp. 309
Segmenting the data warehousep. 310
Staging the warehouse implementationp. 311
Kick-starting the implementation processp. 316
Coordinating the data warehouse implementation processp. 329
Critical success factorsp. 331
Conclusionsp. 332
Justifying the warehousep. 335
The traditional justification approachp. 336
Beyond cost avoidancep. 339
A new basis for competitivenessp. 340
Changing management structuresp. 344
The automation of marketingp. 345
Data warehouse costsp. 347
Conclusionsp. 349
Organizational implications of data warehousingp. 351
From planning to pilotp. 352
From initiation to roll-outp. 358
Conclusionsp. 362
Physical structure of the data warehousep. 363
The data warehouse environment--centralized versus distributedp. 364
Aligning the data warehouse with the organizational structurep. 371
Subsetting the BDWp. 377
Conclusionsp. 378
Data warehouse managementp. 379
Replication administrationp. 380
From administration to runtimep. 389
Process managementp. 392
Data transferp. 394
Other database support functionsp. 397
Conclusionsp. 399
Looking to the futurep. 401
A single information sourcep. 402
Distributed information availabilityp. 404
Information in a business contextp. 406
Automated information deliveryp. 407
Information quality and ownershipp. 408
Concluding remarksp. 409
Referencesp. 411
Indexp. 417
Table of Contents provided by Syndetics. All Rights Reserved.

This information is provided by a service that aggregates data from review sources and other sources that are often consulted by libraries, and readers. The University does not edit this information and merely includes it as a convenience for users. It does not warrant that reviews are accurate. As with any review users should approach reviews critically and where deemed necessary should consult multiple review sources. Any concerns or questions about particular reviews should be directed to the reviewer and/or publisher.

  link to old catalogue

Report a problem