Client-Server Models

In the ›rst chapter of this book, we considered client-server architecture to represent a single model that varied by the manner by which data ³owed between each computer. If data ³owed directly from client to server, the architecture could be represented as what is referred to as a two-tier architecture, with the client considered to represent the ›rst tier, while the server is considered to represent the second tier. If data ³owed from client to server and, depending upon the request, then ³owed to another server, we could refer to the architecture as being a three-tier architecture. As a review, in a two-tier architecture, the user interface is typically located in the user’s desktop, while the database being accessed is located on a server that provides services to many clients. In a three-tier (also referred to as multitier) architecture, a middle layer was added between the client and the database to be accessed. e middle layer can queue requests, execute applications, provide scheduling, and even prioritize work in progress. When considering trade-o‘s between a two-tier and a multitier architecture, it is important to note that the latter will always increase the data ³ow on a LAN, and this increase needs to be considered, especially if the network is approaching congestion prior to implementing a multitier solution to a database retrieval problem. In addition, because a number of di‘erent types of software products can reside at each tier, this can result in a series of di‘erent client-server models, which is the topic of this chapter.