How to Design a Database Application
Database applications are extremely useful in a wide variety of contexts. Creating a database application is relatively straightforward, but does normally require learning one or more technologies. Which technologies you use will depend partly on how you intend to deploy your database application. Taking the time to come up with a strong database design before you start developing your application is always ideal. A well designed database will form an effective, efficient basis for your application, as well as being robust and easy to administer. Whether your database application is for the Web or for a Desktop application, there are a series of common steps that will likely be involved.
Other People Are Reading
Instructions
-
-
1
Design your database starting with a conceptual model. Spending time creating a good database design is something that can have a huge impact on how effective, efficient and useful an application turns out to be. There are many different techniques that can be used, including Entity Relationship diagrams. For any relational database system, you should start by deciding on a set of data Objects. Each Object will correspond to a table in your database, and should have a set of attributes, each of which will be a column in that Object's table. The Relationships between Objects can be represented in the database using Foreign Keys. Each table should also have a Primary Key making every record in it unique. (See References 1)
-
2
Build your database according to your abstract design. If your database is going to be used on the Web, there may be automated tools provided by your Web host that you can use to create it. In many cases, you will need to create the tables in your database by defining them in SQL statements, and then running these within your Database Management System. The SQL used to create tables should define the table names, their column names, and the data types to be used within the columns. (See References 2)
-
-
3
Define the features of your database. When you create the tables in your database, you can also define such characteristics as Primary Keys and whether columns can have "null" entries. You can also use optional features such as Auto Increments to assign the next in a series of numbers each time a new entry is added to a table. This is commonly used for Primary Key columns. Insert some data into your database, even if you can only insert test data to begin with. (See References 3)
-
4
Connect to your database and define the set of queries that will be used. For most database applications, there will be a set of actions that users are allowed to carry out on the data. Some Database Management Systems will allow you to define exactly what privileges a given user has. A common approach is to define one or more abstract users of the database, for example "administrator" or "manager" depending on the actions that particular users need to perform. Within your application code, you should define the set of queries that can be performed, which will include not only viewing, but also updating and inserting data. (See References 4)
-
5
Design and build your database application interface. How you build the interface for your application will depend on where you plan to deploy it. For the Web, you will typically need to use HTML or XML in some form, normally in conjunction with Client Side technologies such as JavaScript and CSS for styling. For a desktop application, you may use a language such as Java to both connect to your Database Management System and present the Client interface to users. Your interface will naturally have to enable users to carry out the functions that your set of queries are intended for. (See References 5)
-
1
Tips & Warnings
Make sure your database application has been tested extensively before it goes live. This is one of the main reasons it's a good idea to use test data while you're developing the application.
Be wary of any database design that involves duplication of data. An efficient database design is one in which there is only "one point of change" for any given item of data.
References
Resources
- Photo Credit black database image by .shock from Fotolia.com
