Topological Analysis and Visualization of Network Performance
SLAC Contact: Dr. Yee, Jerrod Williams, Connie, Adnan Iqbal
Domain and Requirements
The project is to contribute to production software packages that will facilitate the identification of network performance scheduling, monitoring, event detection and visualization. In particular, PathNeck (an opensource tool developed jointly by AT&T and CMU ) has been identified as a useful network monitoring tool that we wish to implement and extrapolate information from. It combines traceroute information with throughput information from packet train deductions.
The project consists of following components:
• Identification of the intrusiveness and accuracy of the PathNeck tool. Comparison of pathneck compared to other tools, in particular traceroute and pathchar derived tools.
• The integration of the PathNeck tool into the IEPM-BW monitoring suite: this involves the development and of the MySQL database tables for PathNeck data and adding the PathNeck tool into the IEPM-BW suite in order to automatically schedule and run the tool.
• Create web based application that will represent textually as well graphical representation of the PathNeck information via MySQL retrieval mechanisms to tabulate PathNeck data in summarized and understandable formats.
• The implementation of classes that will integrate with the existing visualization architecture to retrieve information from the MySQL PathNeck database tables to graphically represent the PathNeck data (for example using Graphviz ).
• Design and implement intelligent algorithms to detect bottleneck and congestion points across networks using the above components and develop example case studies of performance across various networks.
• Advanced candidates will also design and develop modular architectures that will facilitate the automatic detection of changes in routes and/or throughput data (event detection) - using the PathNeck database tables as an example.
1. Existing tools like tracepath, tracenal, pathneck, pathchar produce their results in text mode, which are good for experts to use for on-demand review of a single route, but for the end user they are hard to understand and do not give an overall network view. Therefore this tool will try to bridge this gap and provide a deployable toolkit to meet this need..
2. This tool can be very beneficial to the Long distance/backbone providers to detect the network flaps that occur on the international routes or with other carriers.
3. The toolkit will be deployed to support data-intensive science at SLAC, CERN and other LHC sites. In addition we will explore deploying the tool at ESnet and NTC/Pakistan.
4. Contribution to the open source community.
5. Close interaction with researchers and developers from SLAC.
6. Exposure to network performance measurement tools and techniques
7. Expertise in network performance measurement and visualization techniques.