3 Tier Client/Server Architecture

Most sophisticated Web based applications which involve data entry are based on a 3 tier client server architecture. The 3 tiers are

• the Client (Web Browser)
• the Web Server/Application Server
• the Database Server

Data validation can happen in any or all of these 3 tiers depending on the needs of the application. There are pros and cons to validating data in each tier as described in the following sections.

The Client

The client is the Web browser that runs at the Users PC or Workstation. Validating data input at the Client provides the most immediate feedback to the User and is the main focus of this project. JavaScripts, Java Applets and ActiveX controls are client side components which can be used to accomplish this validation.

Advantages:

• Provides most immediate feedback to the User
• Avoids additional error screens and pages
• Mimics traditional dedicated applications

Disadvantages:

• Slower to load page
• Data validation rules are separated from the database

The Web/Application Server

The center tier in the architecture is the Web/Application server. This includes the Web Server and the application specific code that gets executed by the Web server. This could execute in the same process through vendor specific interfaces or in separate processes via CGI. So the application server can be in many forms including Perl scripts, Java programs, C/C++ programs, active server pages, etc. In fact, it could be a combination of these things.

Advantages:

• Can be implemented in many languages
• Can sometimes reuse existing code if porting a legacy application
• More responsive to the user than Database validation
• Can detect multiple errors at the same time

Disadvantages:

• User most "submit" the data before it is validated.
• Data checking is done separate from data storage.

The Database Server

The 3rd tier is the Database Server which is usually located on a different system than the Web/Application server. This is because the data is often used my more than one application or process. Even if it is located on the same system, it is still an architecturally separate tier. Typical database servers used today are Microsoft's SQL Server, Oracle, DB2, Sybase, etc.

Advantages:

• Single point of validation if many different applications use the same database.
• Data consistency rules stored with the actual data.

Disadvantages:

• Furthest point of validation from the User. (Slowest response.)
• Sometime difficult to implement complex cross-field validation.
• Typically reports only one error at a time. A User could enter several invalid fields and only be notified of one error at a time resulting in many round-trip attempts to "submit" the data.