Analysis of Academic Web Server Traffic and Workload Characterization for Performance Evaluation

Abstract

The web 3.0 considers the latest version of world wide web (www) is increasing rapidly nowadays, as the interest of users in online banking, commerce, business, and online learning about the web. Mostly, the requests made by the users have given a high response time due to the large traffic on the web-server. The services provided by the web-server must be available and maintainable, so there is an extremely need to analysis of web-server’s performance. The proposed study addressed the problem to analysis of real-world web-server workload characterization. Initially, the dataset was prepared for the analysis technique with the eliminations of anomalies, noises and null values. Next, the workload parameters have identified and defined for exploration. After that the techniques is used to group similar data according to their characteristics and further data analysis has performed to extract meaningful insights. Based on the workload characterization insights, performance evaluation has measured using regression analysis and measure of dispersion.