PerfSONAR data visualization

• Main Author: Mohamed Nazri Mohamed Yahya
• Other Authors: Lee Bu Sung
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/69158
• Institution: Nanyang Technological University

Networks are an integral part of society. Other than being the backbone of the internet, it is also an important educational tool to share resources amongst educational institutions. As networks continue to grow, it becomes unpredictable and difficult to maintain or troubleshoot in an event of failures. With that, network measurement tools are becoming increasingly important. One of such tools is PerfSONAR, a tool developed by Internet2, ESnet, Indiana University and GEANT through an international collaboration. PerfSONAR is an open source network measurement tool having the capability to measure a wide variety of network metrics. Such tools empower network engineers as it provides valuable insights on the networks, and to assist in managing end user expectations. With the tool becoming dominant in the networking space, there is a lack of visualization to display the data in an informative manner. PerfSONAR Data Visualization aims to create a web based application for the network engineers to observe the network from an overview perspective. Current visualizations built for PerfSONAR are only able to display from the perspective of a single instance of PerfSONAR. The web based application also provides the feature to highlight networks with abnormal network measurements which are considered as being above the threshold defined by the Upper Fence. Beyond the features, PefSONAR Data Visualization looks at adopting proper frameworks for both Front-End and Back-End development to support code maintainability and extensibility for future enhancements.