Academic journal article Issues in Informing Science & Information Technology

Redesign of Stand-Alone Applications into Thin-Client/server Architecture

Academic journal article Issues in Informing Science & Information Technology

Redesign of Stand-Alone Applications into Thin-Client/server Architecture

Article excerpt

Introduction

Since the early eighties, at the time that personal computers where introduced in a computer industry that was dominated by mainframe computers, the technology of client/server computing started to emerge for sharing data and applications across a network of computers. During the evolution of client/server computing many technologies are developed to support this architecture and client/ server still is a dominant architecture in computer industry. As compared to stand-alone applications and mainframe systems, advantages of client/server computing are flexibility, interoperability, performance, robustness, distribution and scalability.

The aim of this paper is to evaluate possibilities and propose a strategy for redesigning a standalone application to client/server architecture in order to adopt the architecture's advantages. The necessity for this project has come from a recently developed Point-Of-Sale (POS) application. This application was originally designed to operate in a stand-alone environment to ensure the application's availability, even in case of infrastructure failure. Although this design choice has resulted in a robust solution, the application requires a significant amount of hardware resources in order to obtain reasonable performance. These hardware requirements limit the usage of the application to high-end desktop computers while customers in retail business demand a more flexible and wider range of usage. Examples of customer demands vary from usage of network computers in order to reduce the costs of infrastructure, to usage of Tablet PC and Pocket PC as mobile POS units for assistance in peak seasons.

The flexibility of the POS application can be accomplished when moving to thin-client/server architecture as this will reduce front-end hardware requirements and opens new opportunities. The advantages of a client/server POS application:

* Less resources of POS hardware required;

* Flexibility in client applications. For example pre-scan units where customers can scan their products themselves instead of the cashier;

* All POS units have identical product and price information because they share the same information resource;

* Easier and faster replacement of POS units because (product) information does not have to be downloaded to the local machine;

* Additional POS units can easily be added in peak seasons.

The obvious disadvantage of the client/server solution is the dependency of the network infrastructure. To achieve equal availability in the new application more effort and complexity is required in infrastructure by adding redundant components.

The goal of this paper is to examine client/server architectures and possible solutions for redesign of existing applications to this architecture, and to propose a strategy to redesign the POS application. For evaluation of the proposed strategy, the feasibility of a client/server POS and the technical aspects of the implementation, a proto-type application is developed. This proto-type application can be used in future as a model for implementation of the current application.

Redesign Approaches

Reengineering and redesign of applications has become a very important discipline in software engineering, especially in the area of legacy (mainframe) applications. Although the software application in this project can not be considered a real legacy application, it only exists for three years, the years of experience in reengineering and redesigning legacy applications is an important source of information for this project.

Reasons for reengineering and redesign techniques to become this important is because software tends to never reach a point of being finished, and the operational lifetime is quite unpredictable at time of design, implementation and deployment. Some applications will not exceed a few years of operation while other applications will still be business-critical for an organization after ten or even twenty years. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.