Prevalent Database Models in Software

Creation UPC Code in Software Prevalent Database Models

A database model is essentially the architecture that the DBMS uses to store objects within the database and relate them to one another The most prevalent of these models are presented here in the order of their evolution A brief history of relational databases appears in the next section to help put things in a chronological perspective

Flat files are ordinary operating system files, in that records in a file contain no information to communicate the file structure or any relationship among the records to the application that uses the file Any information about the structure or meaning of the data in the file must be included in each application that uses the file or must be known to each human who reads the file In essence, flat files are not databases at all because they do not meet any of the criteria previously discussed However, it is important that you understand them for two reasons: First, flat files are often used to store database information In this case, the operating system is still unaware of the contents and structure of the files, but the DBMS has metadata that allows it to translate between the flat files in the physical layer

and the database structures in the logical layer Metadata, which literally means data about data, is the term used for the information that the database stores in its catalog to describe the data stored in the database and the relationships among the data The metadata for a customer, for example, might include all the data items collected about the customer (such as name, address, and account status), along with the length, minimum and maximum data values, and a brief description of each data item Second, flat files existed before databases, and the earliest database systems evolved from the flat file systems that preceded them Figure 1-2 shows a sample flat file system, a subset of the data for fictional company Northwind Traders, a supplier of international food items (and a Microsoft sample database) Keep in mind that the column titles (Customer ID, Company Name, and so on) are included for illustration purposes only only the data records would be stored in the actual files Customer data is stored in a Customer file, with each record representing a Northwind customer Each employee of Northwind has a record in the Employee file, and each product sold by Northwind has a record in the Product file Order data (orders placed with Northwind by its customers) is stored in two other flat files The Order file contains one record for each customer order with data about the orders, such as the customer ID of the customer who placed the order and the name of the employee who accepted the order from the customer The Order Detail file contains one record for each line item on an order (an order can contain multiple line items, one for each product ordered), including data such as the unit price and quantity An application program is a unit of computer program logic that performs a particular function within an application system Northwind Traders has an application program that prints out a listing of all the orders This application must correlate the data between the five files by reading an order and performing the following steps:

1 Use the customer ID to find the name of the customer in the Customer file 2 Use the employee ID to find the name of the related employee in the Employee file 3 Use the order ID to find the corresponding line items in the Order Detail file 4 For each line item, use the product ID to find the corresponding product name in the

Product file This is rather complicated given that we are just trying to print a simple listing of all the orders, yet this is the best possible data design for a flat file system One alternative design would be to combine all the information into a single data file with all the data about the customer, employee, and order combined into a single record for each order Although this would greatly simplify data retrieval, consider the