Networking Research Lab
Computer Science DepartmentThe University of Memphis

Photo: The University's FedEx Institute of Technology, which houses several Computer Science facilities



CRI Collaborative Research: Building the Next-Generation Global Routing Monitoring Systems

The project is a joint 3-year development effort among Advanced Network Technology Center (ANTC) at the University of Oregon, Colorado State University, University of Memphis, and University of California at Los Angeles to build the next generation of the RouteViews system. Our project will enable this global routing monitoring system to provide high-quality data and realtime access to the research and education community as well as the network operations community. Understanding global routing is critically important for current Internet research. In particular, it will guide the development of new routing models and paradigms. The existing Internet relies on the BGP routing protocol, but limitations in the existing system are also well known: security issues and convergence problems. Although it is not clear how a successful design of next generation global routing protocol would look like at this time, one thing is clear: the data that can help us understand the current global routing system is essential to a wide range of network routing research e Understanding the current system requires access to the operational Internet which can be a big barrier to research community. Commercial Internet service providers can be reluctant to widely share their operational data. While these barriers can (and have for a small group of researchers) be overcome by determined e is obtaining global scale data. To truly understand and properly analyze global routing data, one needs data from a wide range of sites with different types (tiers) of ISPs. Fortunately global routing monitoring projects such as Oregon RouteViews have been providing such essential data to both operations and research community. RouteViews collects BGP routing data from a number routers around the world and makes this data freely accessible.

RouteViews has a proven track record in benefiting the research community as well as the operations community, but it was largely put together in an ad hoc way to meet operational needs. Extensive use of RouteViews data has exposed a number of serious limitations in the current system. These limitations create obstacles in using the data for research and education, thus blocking the way to meet new operational as well as research needs. A new development to help RouteViews overcome these limitations is needed to prevent this community resource from becoming outdated and obsolete.

We believe that time is overdue to re-engineer RouteViews, a critically important global routing collection system, to produce high quality of data and meet today and future demands. We formulated these demands into the following list of goals of the new RouteViews.

  • Data Completeness
  • Data Integrity
  • High Accessibility
  • Scalability and Extensibility

To achieve the above goals, we have decomposed the engineering challenges into the following three ma jor tasks. First, we will develop modern data collection software to gather the data from production routers. Instead of providing the routing functionality as the current routing software suites, the data collection software focuses on maintaining peering sessions and collecting BGP updates. Second, we will design and develop a new data storage system and add new meta-data to facilitate e new access system that enables real-time data access services. Each of the three components of the system has a specific functionality. The first component supplies the data collected from production routers to the other components, while the other two components mainly interact with users. The data storage system provides archived data and new access system serves newly collected data to the users. Collectively, they form a reliable and scalable global routing monitoring system.


Infrastructure of the monitoring system.