Transforming the Logical Model to a Physical Database Design

A logical data model describes the data and the business rules pertaining to that data without regard to how the data will be physically stored. The physical data model describes how that model will be implemented. It shows the table structures, data types, referential integrity, and more. Denormalization may occur as well and partitioning. It may look very different from the logical model. And there could be differences based on which RDBMS is to be used.

This session will begin with a logical model and look at the various aspects that should be considered when transforming the logical model into a physical database design. The decisions you make at this point can have an impact on performance, concurrency and data integrity. This session will include:

discussions of database volume
database access and usage
data types
key structures
constraints and
referential integrity
denormalization and
index strategy

from a DBA's perspective.

Christine Kress is a Sr. Data Analyst / Modeler at Lowe's Companies, Inc. She has seven years of experience as a Data Analyst backed by 17 years as a database administrator. She has worked primarily with DB2 z/OS, but also has experience with SQL Server, IMS and EDMS. She brings her many years of DB2 DBA experience to discuss the options to consider when taking a logical data model to a physical design. Chris is a member of DAMA Carolinas.