OVERVIEW

There is a plethora of material on normalization, mostly against it and outright wrong, but despite its having been thoroughly debunked, confusion and poor understanding still reign in the IT industry and, sadly, in academia too. Most of the material accessible to the practitioner is in what we call “cookbook” form—“recipes” to be followed mechanically without understanding why. This contributes to the poor understanding of normalization, its practical purposes and particularly the severe costs of “denormalization” that are almost completely ignored.

Instead of focusing on the real solution to performance maximization—better implementations of truly relational DBMSs (TRDBMS) and physical design—the industry distracts practitioners who, like vendors and the trade media, are mired in the logical-physical confusion, by encouraging or compelling them to “denormalize for performance”, to the consequences of which all are oblivious.

Correct business modeling and logical design yield databases that adhere to the Principle of Full Normalization (POFN) and the Principle of Orthogonal Design (POOD) and benefit from their practical advantages. A normalization procedure is required if and only if database design is flawed—it is, in other words, a design repair endeavor.

OBJECTIVES

The focus of this seminar is on understanding normalization and its practical benefits, and undernormalization and its costs. It

Ø Explains normal forms

Ø Documents all costs of undernormalization

Ø Exposes “denormalization for performance” as an illusion and explains why it is sometimes induced by DBMS products

Ø Debunks some prevalent misconceptions

OUTLINE

· Introduction

· Understanding Normalization

· The Full Price of Undernormalization

· Further Normalization

· "The Whole Key" & Second Normal Form

· "Nothing But the Key" & Third Normal Form

· General Dependencies

· Multivalued Dependencies & Fourth Normal Form

· Join Dependencies and Fifth Normal Form

· Two Database Design Principles

· "Denormalization for Performance"

· The Logical-Physical Confusion

· Database Bias

· Redundancy Control

· The Real Problem and Solution

· Some Fallacies Debunked

· Conclusions and Recommendations

AUDIENCE

Anybody involved in data management, technical and not technical. Some data management background may or may not be helpful. The target audience includes (but is not limited to):

§ DBMS designers, implementers, and other vendor personnel

§ Database consultants

§ Data and database administrators

§ Product evaluators, acquirers and deployers

§ IT managers

§ Information modelers and database designers

§ Application developers and deployers

§ Data warehouse implementers

§ Members of the trade media covering data management

§ Academics specializing in data management topics

§ Students, graduate and undergraduate

DOCUMENTATION

Workbook containing the instructor’s slides.

INSTRUCTOR

Fabian Pascal has a national and international reputation as an independent technology analyst, consultant, author, and lecturer of seminars, specializing in data management. He was affiliated with Codd & Date and for more than 30 years held various analytical and management positions in the private and public sectors, has taught and lectured at the business and academic levels, and advised vendor and user organizations on data management technology, strategy and implementation. Clients include IBM, Census Bureau, CIA, Apple, Borland, Cognos, UCSF, and IRS. He is founder, editor and publisher of DATABASE DEBUNKINGS, a web site dedicated to dispelling persistent fallacies, myths and misconceptions prevalent in the IT industry. He publishes PRACTICAL DATABASE FOUNDATIONS series of papers dedicated to explaining the logic foundations of database management to IT practitioners. Author of three books, he has published extensively in most trade publications, including DM Review, Database Programming and Design, DBMS, Byte, Infoworld, Computerworld and various web columns.

Updated 1/26/11