Chapter 8 Summary Dbms

Chapter 8 summary

It’s not easy to build a database correctly the first time, because even if we obtain all users’ requirements and build a correct data model, the transformation of that data model into a correct database design is difficult. Information Systems influence organizations and that organizations influence information systems, and in many ways they create each other. We want to make minimum impact on existing data, when we try and change a database that has data. A correlated subquery is a type of subquery in which an element in the subquery refers to an element in the containing query.

The EXITS keyword will be true if any row in the subquery meets the condition. The SQL NOT EXISTS keyword will be true only if all rows in the subquery fail the condition. The double use of NOT EXISTS can be used to find rows that do not match a condition. Reverse engineering is the process of reading the structure of an existing database and creating a reverse engineered data model from that schema. A reverse engineered data model is the structure that results from reverse engineering; it’s not really a data model, because it includes physical structures such as intersection tables. It is a median between a data model and a relational database design.

Dependency graphs are not graphical displays like bar charts, rather they are diagrams tha consist of nodes and lines that connect those nodes. Dependency graphs should be used to understand the consequences of changing a table name. changing minimum cardinalities on the parent side of a relationship can be down by altering the constraint on the foreign key from NULL to Not NULL, but changing it on the child ‘s side of a relationship can be done by adding or dropping triggers that will enforce that constraint. Reducing cardinalities is easy, bust those changes can result in data loss. Forward engineering is the process of applying data model changes to an existing database.