Customer Reviews

The Data Warehouse Toolkit: The Complete Guide to Dimensional Modeling (Second Edition)

5 star:		(28)
4 star:		(7)
3 star:		(2)
2 star:		(0)
1 star:		(3)

 

 

There are a lot of data warehousing books out there that try to answer the question: 'Why'? Why data warehouses are needed to help businesses make better decisions - why the OLTP systems that run the business can't do this - and sometimes even why businesses ought to invest in data warehouses. These books were terrifically useful to us years ago, when we needed help (and scholarly footnotes) in our data warehouse project proposals. This book is not one of those - it is all about:

How

How to actually design and build a repository that will deliver real value to real people. In this reviewer's opinion, Ralph Kimball's many contributions related to the 'how' of data warehousing stand alone.

An engineer wishing to jump-start his or her data warehouse education would need to read Ralph's Data Warehouse Toolkit first edition, his Data Webhouse Toolkit... a bunch of "Data Warehouse Designer" Intelligence Enterprise magazine articles... AND lurk on the Data Warehousing List Server...for a few years (all terrific resources - by the way) - in order to stockpile the knowledge that is crisply presented here.

No shortcuts taken by the authors that I can spot: all of the toughest dimensional design issues that I've tripped on - and that I can remember surfacing on in discussion groups over the past few years - are addressed in this significantly updated text. Not all of the solutions are 'pretty' - but it is clear that they thoughtfully address the problem. This approach, in my opinion, instills student confidence - and lets us know that we are getting sound instruction - not dogma.

The authors have been listening to and addressing the data warehouse community's 'pain' through periodicals and posts for years - but this book pulls these point solutions together very nicely. I learned a surprising number of really useful new techniques, and was genuinely enlightened by the 'Present Imperatives and Future Outlook' section.

As in the first edition, there is minimal philosophical lecturing, and zero religion. Instead, we get generous helpings of real-world case studies - aptly applied to progressively more advanced series of design concepts.

This style absolutely works for me. And I suspect that engineering mindsets typical of the folks that build these things will likely agree. In short, the Data Warehouse Toolkit Second Edition will significantly lighten the load of books that I carry between data warehouse engagements.

Jim Stagnitto
Llumino, Inc.
www.llumino.com

If you want to understand data warehouse design either as user, architect or developer, you need to read this book cover to cover.

Things I like about this book:

* Coverage of all core principles in dimensional data modeling using examples. Ralph does not just lecture to you -- he shows you how to put it into practice

* Coverage of a vast variety of domains. This alone makes the book a must-read

* Recap of major principles at the end of the book to bring it all together

* Excellent writing -- Ralph does not treat you like a dummy; neither does he assume that you have an IQ north of 200

* When you purchase this book, you are in effect purchasing a sliver of the combined knowledge of both authors in the data warehousing field. Highly recommended

I implemented a data warehouse using some of these principles back in 1999. The project was a resounding success and is the most popular application in the financial services firm that I implemented it in. (Infact when I lost my job at an Internet company, they immediately offered me a job based on this implementation). The only sad part to the whole story is that we made a few mistakes in implementation that are now very difficult to correct because the data warehouse has become core to the business -- we have too many end-user applications riding on it!

After six years of creating data warehouse applications, making a plethora of mistakes and learning stuff the hard way, I wish I had had this book at the start! Every other page offers a solution to some problem or other that I have had. In the project I am just starting I am facing new challenges and am finding help with them as well. The best part is how solutions I used in the past which were appropriate for those problems are contrasted with solutions for problems like the ones I am facing now. Almost as bad as solving a problem the wrong way (or overlooking it entirely) is reusing an old solution that does not fit the new problem. This book clearly spells out when each solution is appropriate. I can not speak too highly about how useful this book will be for you!

Data Warehousing is more of an art than a science - but then again what isn't?

There isn't a standard blueprint that can come close to solving most data issues. Data Warehousing (DW) involves constant tweaking and the goal of good DW project management is minimizing the associated operational cost.

I have been a fan of Ralph Kimball as he writes as a person who has been through many implementations. With Mr. Kimball there isn't a miracle cure being touted - stay away from publications that claim such a cure.

Mr. Kimball approached the subject with good advices and encourages the readers to watch out for the pitfalls and follow best-practices in design implementation. It is similar to working with a well experienced supervisor.

The core to successful DW implementations is - LISTENING. Listening to the users on their needs and gauging the software resources available at your disposal.

Trade-offs in design versus cost/performance are a must. You will never have all the resources you need to implement the DW of your dreams. And if you did, chances are very high that once the DW is ready for use the business cases have changed making the design redundant.

Mr. Kimball will help in passing these information and much more. It also goes in good technical detail for suggested modeling of data.

I hope this review is helpful, please let me know if you have any questions or suggestions.

I was undecided if I should buy this book after having read the first edition, but I'm happy I did.
The second edition updates many of the concepts contained in the first and includes some new chapters on hot topics like CRM and Telecommunications (which is the most important sector for dw at least here in Italy where I live).
I think that Kimball books are everything that's needed to design good, robust and flexible data warehouses, and this book maintains his high quality standards.

This book deserves merely a one sentence review.
"Simply awesome" . As one of my friends had to say,
"We depend on Ralph Kimball for our livelihood"!

This IS the book on Data warehousing. Follow this book
and you will never fail. I have had solutions to all
of my design issues from this one book.

A must to every IT professional's personal library,
not necessarily restricted to DW professionals

Excellent work. When I first saw it I thought it was mere rewrite with simple correction of the first edition that I had bought in 1996. Thankfully I still sit down with the second edition and I realized how much additional insight I would have missed had I not read the second edition. It is very well written, thought provoking and easy to follow and yet conveys many complex concepts.

Congratulations to Ralph and Margy.

In my mind this book is defintely the most definitive guide to dimensional modeling. Although there are some chapters (the last two that deals with process and other issues, the biggest part of the book is solely about dimensional modeling. Yes the first chapter deals with what it is (in general), why we would like to use dimensional modeling and some myths and caveats. However, from chapter 2 through 12, the book is hand-on.

Chapters 2 - 5 is for me the most impressive chapters of the book as it takes several business processes that is common in many businesses and discuss the design of the star cshemas in great depth. However, nowhere in the book is any design upheld as a panacea -- everywhere the book discusses at great length the possible problems and the design decisions to be made. It thus emphasizes that there is no "one-size fits all" in dimensioanl modeling.

Chapters 6, 7 and 8 deals with CRM, Accounting and HR respectively, while chapters 9 - 14 deals with different industries and their peculiar problems. The industrues are: Telecomms and utilities, Transportation, Education, Healthcare and e-commerce. Each chapter introduces new design guidelines and new problem areas anbd their possible solutions. Chapter 15 deals with the Insurance industry and essentially serves as a nicesummary of the proposed techniques as most are emplyed in the Insurance case study.

If you do not have a fairly good understanding of dimensional modeling after reading this text I would be very surprised.

However, it is more than just a read to know about dimensional modeling. The way the case studies are handled and the problems identified and discussed also makes it a true reference book. I think evrybody involved in dimensional modeling should have one on his/her desk.

This book delivers exactly what it says. Except that word "toolkit" in its title - you'd better think about this book being a single tool, not a whole toolkit. Ralph Kimball actually has a whole lot of books on data warehousing published, this is one of them, a tool in the toolkit. This one seems like a good starting point to the entire series, and it only shows a single facet - the dimensional modeling.

The book explains the basic principles of creating dimensions and fact tables in a data warehouse (assuming a relational star schema), and then dedicates a chapter per industry to show how those principles apply to sales, order management, CRM, accounting, human resources, financial services, telecoms, logistics, education, health care, e-commerce, insurance etc. Each one appears to be significantly different from the others.

There is a couple of teaser chapters starting with "we have that other book covering this, but will brief you out". Nice and makes you want to read the other books too.

The book also includes guidelines to the warehouse building process, in terms like "know your business sponsor", "talk to your users" and so on. Difficult to say what it has to do with dimensional modeling, perhaps it's included in all the books in the series.

There is no word on software, hardware, physical architecture, tuning or performance in this book. It is a textbook in dimensional modeling, period.

The book is written clearly, has a handful of simple and uniform diagrams and is easy to follow. It only leaves you wondering just how exactly large is the whole data warehouse area, how many pieces you need to collect yet.

Recommended.

This book is a great primer on data warehouses: what they are for, how data should be organized in the warehouse, and what you can do with it. There's no code or programming - just a solid explanation of the concepts along with many good examples.

This book was perfect for me. I have data that I needed to analyze; I understood normalized relational databases and pivot tables; I'd heard the jargon such as "star schema" and "snowflake schema" but didn't know exactly what they meant. This book put all the concepts together and moved my level of understanding up to a new level. I am not building the data warehouse myself but now I can talk the talk with those who are.

I had been using pivot tables in Excel to do some analysis, but had been frustrated at times by not being able to get things to work out as I wanted. After reading this book, I understand why - it was because I had not organized the underlying data correctly. To anyone with a relation database background, the organization of data in a data mart or warehouse is very different and goes against one's instinct to normalize and eliminate redundant data. Getting past that was key to getting my data right, and that's what this book did for me.

However, if you are the person who will actually be building the data warehouse, you will need to go beyond this book to one that shows you how to actually do what you need. For example, this book does not go into "details" like ETL - how you actually get data into the warehouse from the transaction database - and the other practical details you will need to know. This book is also database and tool agnostic - those are all details that you will need to find elsewhere once you understand the concepts here.

My only quibble is that this book is quite pricey for what you get.