Site multihoming is a method by which an Internet end-site, for example an enterprise network, may connect to multiple service providers simultaneously. There are many reasons why multihoming is desirable, e.g. service resilience, network load balancing or provider independence. In the IPv4 Internet, multihoming has been achieved by use of relatively simple techniques, including networks advertising their network prefixes -- whether such prefixes are independent of the Internet Service Providers (ISPs) or not -- to the Internet global routing infrastructure. With the introduction of IPv6 the vast increase in the number of potential site prefixes means that for scalable site multihoming we cannot repeat such IPv4 multihoming practices. Thus new IPv6 multihoming solutions are required. In this paper we present an overview of currently proposed solutions and explore the challenges and motivations of site multihoming. Such a review is timely because multihoming remains the main perceived obstacle to widespread IPv6 deployment in mission-critical environments.

In this paper we introduce a framework for analyzing BGP connectivity, and evaluate a number of new complexity measures for a union of core backbone BGP tables. Sensitive to engineering resource limitations of router memory and CPU cycles, we focus on techniques to estimate redundancy of the merged tables, in particular how many entries are essential for complete and correct routing.

The need to service populations of high diversity in the face of high disparity affects all aspects of network operation: planning, routing, engineering, security, and accounting. We analyze diversity/disparity from the perspective of selecting a boundary between mice and elephants in IP traffic aggregated by route, e.g., destination AS. Our goal is to find a concise quantifier of size disparity for IP addresses, prefixes, policy atoms and ASes, similar to the oft-quoted 80/20 split (e.g., 80% of volume in 20% of sources). We define crossover as the fraction c of total volume contributed by a complementary fraction 1-c of large objects. Studying sources and sinks...

The current Internet routing landscape presents a number of challenges, especially in the configuration of routing policies. There have been numerous proposals to tackle the misconfiguration problem: configuration checking, policy language (re)design, and clean-slate routing architecture. In this paper, we present an analysis of routing policies and a misconfiguration detection mechanism. With an operational perspective, we first present a study on the configuration and evolution of routing policies, using data from three different types of networks. Our results show that configurations are changed frequently and mostly incrementally. We found that the most commonly used and changed commands are related to route tagging and filtering, and there are substantial amount of duplication in policy configurations within a network. More interestingly, based on these results, we develop a data mining method to find inconsistencies in a network’s configurations of routing policies. Our method is able to detect local, network-specific rules automatically, and differs from existing approaches that are based on universally applicable rules. In our evaluation, we found 30 confirmed errors and 29 warnings in three networks. More than half of the errors are related to route tagging. Our findings show that the next generation configuration language and routing platform should be sufficiently flexible to allow a network to express and frequently modify its route tagging, yet restrictive enough as this aspect is often misconfigured. 1

Researchers investigating topics such as performance, stability, and growth of the Internet often turn to BGP routing tables to obtain Internet topology. BGP routing tables provide a mapping from address prefixes to autonomous system (AS) paths. Our study, based on hundreds of thousands of traceroutes from three locations worldwide, categorizes di#erences between AS paths obtained from BGP routing tables and AS paths derived from traceroute paths. We find much of the disparity results from exchange points ASes (which rarely appear in BGP paths) and by groups of ASes under the same ownership. We introduce a new AS relationship, common ownership, that reflects the complexities of real-world business relationships and practices. We conjecture that the observed di#erence in size between the cores of an AS graph derived from BGP and an AS graph derived from traceroute is due to the visibility of peering at exchange points in traceroute paths.