Up to this point, the focus has been on systems analysis.
Eventually, analysis will give way to design. While analysis was focused on answering the
question “What must the system do?”, physical design is concerned with the question
“How must the system do it?” In other words, the technical concerns are addressed during systems
design. |
Since an information system is concerned with data in some form, this data is usually stored in a physical structure
known as a database. A database can be a single text file, or it may involve a series of interrelated files, typically organized in the form of a
relational database. Microsoft Access is a common low-cost relational database used by many businesses, while Microsoft
SQL and Oracle are higher-cost relational databases capable of storing more data for more users more reliably.
Whatever the database technology, an information system design will likely include a physical database design, whether
this be a more structured design for larger projects or a “quick and dirty,” less formal design for smaller projects.
This physical database design will relate to the logical data model produced during analysis, but while the
logical data model may not have been normalized, the complete physical database model will need to be normalized, usually to
third normal form, essentially assuring that each table in the schema is appropriately indexed, entities are
nonredundant, and the database will easily allow for future growth. Additionally, the physical model, which can be
thought of as a complete blueprint for data storage, will be more detailed in that fields, field properties, indexes,
primary and foreign keys, and referential integrity will all be specified. Indicating these details is the following
simple physical database design.
|
