Developing WAN Optimization Model to improve the Performance of Business Critical Applications: the Case of UNECA
No Thumbnail Available
Date
2013-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Wide Area Networks (WANs) have natural characteristics such as high latency, congestion, low bandwidth and high packet loss rate. Due to these factors, the performance of enterprise applications can quickly become degraded when they are being accessed and used over the WAN link. In addition, most of the existing protocols are designed for Local Area Network (LAN) environments; therefore, several application protocols do not perform well over the WAN.
The purpose of this study is to investigate the impact of various WAN factors, with a view to develop a WAN optimization model that can improve the performance of business-critical applications over the WAN. Previous studies reported that an application optimization solution would be economically feasible and effective, if it is designed from the targeted organization WAN environment perspective. Hence, this study attempts to develop a WAN optimization solution particularly targeted for enterprise WAN environments, which have limited network resources.
To realize the goal of the study, various experiments have been undertaken in cases that are taken from the UNECA WAN environment. The experimentation is conducted through four major phases: in the first phase, application’s traffic flow data is collected using Net-Flow analyzer tool from the UNECA WAN environment. Next, the collected traffic flow data is simulated and experimental environment is prepared using the OPNET Modeler tool. The third phase deals with analysis of the simulated traffic data for investigating critical application performance challenges, followed by experimenting different optimization methods, so that the appropriate technique is selected for each performance bottlenecks. Finally, based on the analysis result a WAN optimization Model is developed.
The developed model is designed to identify the real-time bottleneck status of each WAN link and apply optimization techniques accordingly, which are appropriate for improving the performance of each specific application traffic flow. This study suggests further research works for enhancing the efficiency and effectiveness of the Model.
Description
Keywords
Wide Area Networks, (WANs)