Customer Reviews

Oracle DBA Guide to Data Warehousing and Star Schemas

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

 

 

The book offers a number of detailed tuning tips and insights from Mr. Scalzo's real world experience. That being said, the tone of the writing is really annoying. Mr. Scalzo's condescending style is unnecessary and takes away from what could be a very good book. At points, I actually broke out in laughter at his arrogance. However, if you can get past the author's attitude the content of the book seems to be quite helpful.

As an Oracle DBA transitioning from the OLTP world to the warehousing arena, I found this book enormously helpful. In a mere 200 pages, Mr. Scalzo has elaborated on an extensive range of data warehousing topics, including: star-schemas, dimension hierarchies, SQL-tuning, partitioning and parallel loading to name a few. His style is informative and direct; his examples are meaningful and clear!

This is a must-have book for all DBA's serious about designing and managing large-scale, lightning-fast data warehousing systems.

Allow me to bestow some well-deserved praise upon Bert Scalzo's terrific "Oracle DBA Guide to Data Warehousing and Star Schemas". A true gem - I won't go on another Oracle project without it.

What Bert provides here is nothing short of a clear and crisp recipe for success for implementing Oracle-based data warehouses. It fills in a much-needed area of dimensional data warehousing best practices, by describing precisely how to coax the best achieveable Oracle performance from dimensional data models.

I can't tell you how many projects I've been on where I've had to compromise physical data models in order to address perceived "shortcomings" in Oracle's ability to efficiently service dimensional queries. Using Bert's book on my most recent project, we followed his "recipe", and were able to consistently achieve the ideal query optimization plans and aggregate navigation behaviors - simply - without any of the usual hassles that I have (unfairly) come to associate with large scale Oracle data warehousing.

To data warehousing newbies I humbly suggest: pick up any one of Ralph Kimball's terrific texts on data warehouse design, and then if you are rendering it in Oracle, buy this book and follow its advice. The resultant system will be simple, powerful, and fast.

Bravo Bert - a great contribution to the field.

Jim Stagnitto
Llumino, Inc (www.llumino.com)

Scalzo gives a chatty, informal exposition of tuning large data warehouses using Oracle software. He uses a strong conversational tone that makes such issues as star schema query optimisations straightforward to follow, under the assumption that you are already an experienced Oracle database administrator.

The book is independent of Oracle. What is unclear to me is how Oracle would regard this book. In several places, Scalzo describes the appallingly slow performance of Oracle 7 vis-a-vis Oracle 8i and 9i. He basically says that for certain tasks on large data sets, Oracle 7 was badly designed. No one who is currently using Oracle 7 will be thrilled to hear this. (Hee hee.) Cynically, one migh think Oracle 7 was overpromised and it underdelivered in some ways. But positively for Oracle, Scalzo gives typically much better results for 8i and 9i. An inducement to upgrade, if you have not done so already.

The biggest quibble about this book is the chatty style. Doesn't bother me. But it may perturb some of you, if you prefer a more "serious" timbre to the discussion of expensive design and deployment issues. If you can see your way clear past this, then the book may have merit in your DBA job.

Bert has written a very good book that's full of real life empirical facts and experiences, mainly from building a vary vary large DW for 7/11. Bert does especially good job describing I/O, partitioning, spindles. His section on high speed / high volume data loader techniques is also very very good.

I have been leading / managing / directing / coding Oracle software development projects for 15 years for 3 Fortune 100 companies, and am both technical as well as "functional". Highly recommend Bert's book for DBA's, SA's, designers, programmers, technical project managers.

Excellent job of covering various types of databases. If you work with a data warehouse and cannot clearly define what a fact, dimension, or cube is, that's probably why your warehouse acts like a big slow-running database (i.e., it was never designed properly). This book will show you the way.

I bought this one in 2005. At that time I had some experience with MS SQL Server data warehouses but none with Oracle.
I had the fortune to use this book when I was member of the team for the creation of a DWH for a big company, as Oracle is in my opinion the only choice you have for medium to huge DWH.
As a DWH developer and designer I find this book really handy. There are some arguments I don't agree with the author since you cannot speak generally and each implementation has its own issues, but overall the book does help you lots.
I think it is a shame that is such an old version, these days everyone is using 11g or at least 10g but most of the material in the book can work as well with latest versions.
It is somehow difficult at the beggining to figure out how to use all the information in the book but Bert has done a great job describing each problem and pointing out the choices you have to solving them.
I really recommend this one for anyone with some knowledge of Oracle and who is interested in the creation of a Data Warehouse project. The price of the book gets paid off very quickly when you revieve congratulations for the improvement in running queries, and later on more sponsorship from management, which is at the end the ultimate goal of any Data Warehouse.
Thanks Bert, I look forward for your next edition.

Gives the straight scoop on how to make the star transformation work right. Not wicked technical, a book for doers. Non-nonsense, read it in a day, then get the job done. Just what a practitioner needs (at least it is just what I needed). Thanks.

The author definitely has an arrogant style of writing. The book just talks about one way of doing things without being open about other possible better ways of designing and implementing efficient techniques. Oracle features such as partitioning, materialized views, external tables are not done enough justice.

Data warehouse requirements are much more than the two-dimentional approach discussed. There is minimal or no discussion of performance at the system level taking into account infrastruture, architecture, query optimizaion, front-end tools etc.

A very narrow outlook on datawarehouse implementations.

I think this book is exceedingly good.

The relevant, balanced,technical/commercial content backed by code and examples (Particularly the 1 page diagrams) presents Mr Scalzos unquestionable leadership in this field in a simple and unambiguous manner.

Mr Scalzo tells you EXACTLY how to design terabyte dwh databases with as fast as can be achieved response times.

Thank you sir, for telling me the facts.

Further to the above review I would like to add that Mr. Scalzo
provides an example of a materialized view to implement the proposed aggregates on page 167. Perhaps some readers do not understand that an Oracle MV is an actual table, backed by storage.

As for attitude, and style, I personally find this to be totally harmless, illustrative, and for the one or two comments that Mr Scalzo makes, again, in my experience (20 years DBA), to be nevertheless accurate.

The book is a real gem.