"A comprehensive, thoughtful, and detailed book that will be of inestimable value to anyone struggling with the complex details of designing, building, and maintaining an enterprise-wide decision support system. Highly recommended." -Robert S. Craig, Vice President, Application Architectures, Hurwitz Group, Inc.
In his bestselling book, The Data Warehouse Toolkit, Ralph Kimball showed you how to use dimensional modeling to design effective and usable data warehouses. Now, he carries these techniques to the larger issues of delivering complete data marts and data warehouses. Drawing upon their experiences with numerous data warehouse implementations, he and his coauthors show you all the practical details involved in planning, designing, developing, deploying, and growing data warehouses. Important topics include:
* The Business Dimensional Lifecycle(TM) approach to data warehouse project planning and management
* Techniques for gathering requirements more effectively and efficiently
* Advanced dimensional modeling techniques to capture the most complex business rules
* The Data Warehouse Bus Architecture and other approaches for integrating data marts into super-flexible data warehouses
* A framework for creating your technical architecture
* Techniques for minimizing the risks involved with data staging
* Aggregations and other effective ways to boost data warehouse performance
* Cutting-edge, Internet-based data warehouse security techniques
The CD-ROM supplies you with:
* Complete data warehouse project plan tasks and responsibilities
* A set of sample models that demonstrate the Bus Architecture
* Blank versions of the templates and tools described in the book
* Checklists to use at key points in the project
Ralph Kimball
Data-warehouse-Konzept Informatik, EDV / Informatik
(.)Das Buch richtet sich primär an Designer und Informations-Manager im Umfeld von Data-Warehouse-Anwendungen. Diese Zielgruppe erhält Ratschläge und Handlungsanleitungen, um deren Anforderungen optimal zu bewältigen. Dabei ist der ganzheitliche Ansatz zu betonen, der eine evolutionäre Weiterentwicklung besser als singuläre Schnellschüsse unterstützen kann. auch der zugrundeliegende Architekturcharakter trägt zur Festigung dieser Intension bei. Zu knapp werden die Themen Data Mining und Parallelverarbeitung angesprochen; ebenfalls fehlen zur Vertiefung weiterführende Literaturhinweise. Insbesondere SMP und MPP bieten für die praktische Realisierung wichtige Unterstützungshilfen. Wobei zur Verbesserung der Performance die Fähigkeit einiger DBMS Abfragen zu optimieren und parallel zu verarbeiten nützlich ist.IT-Director 10/2000
()