Xenofontas Dimitropoulos FORTH Heraklion, Greece fontas@ics.forth.gr MOTIVATIONA lot of applications depend on reliable and stable Internet connectivity. These characteristics are crucial for missioncritical services such as telemedical applications. An important factor that can affect connection availability is the convergence time of BGP, the de-facto inter-domain routing (IDR) protocol in the Internet. After a routing change, it may take several minutes until the network converges and BGP routing becomes stable again [13]. Kotronis et al. [8,9] propose a novel Internet routing approach based on SDN principles that combines several Autonomous Systems (AS) into groups, called clusters, and introduces a logically centralized routing decision process for the cluster participants. One of the goals of this concept is to stabilize the IDR system and bring down its convergence time. However, testing whether such approaches can improve on BGP problems requires hybrid SDN and BGP experimentation tools that can emulate multiple ASes. Presently, there is a lack of an easy to use public tool for this purpose. This work fills this gap by building a suitable emulation framework and evaluating the effect that a proof-of-concept IDR controller has on IDR convergence time. Categories and Subject Descriptors OBJECTIVESOur primary objective is to support hybrid BGP-SDN experiments with multiple ASes using real router software. This is needed since when deploying a new IDR approach one cannot change the whole infrastructure at once. The framework should take care of configuration management such as IP prefixes and BGP policy templates and the user . should be able to actively control the experiments, e.g., dynamically changing the topology and verifying the effects of changes. Furthermore, it should be possible to easily create topologies based on measured Internet data or theoretical models. This way an experimenter should be able to concentrate more on the experiments and her concept rather than bothering with configuration and management.Our second objective is to demonstrate the effect of centralization on IDR convergence time. We designed and implemented a proof-of-concept IDR SDN controller that exploits centralization to improve IDR convergence time based on the following design goals. First, the controller should inter-operate with legacy BGP routers. Moreover, the cluster network is transparent to the legacy BGP world, therefore ASes within the cluster maintain their AS identity. In addition, we want to support disjoint AS sub-clusters controlled by the same controller, so that an intra-cluster link failure does not isolate the controlled ASes: paths over the legacy Internet could still connect the sub-clusters. HYBRID SDN & BGP EMULATION FRAMEWORKThe framework is based on a slightly modified version of Mininet [10] that supports the functionality of Quagga [3] -a popular BGP software. The topologies can be either artificial or built from the iPlane Inter-PoP links [12] and the CAIDA AS Relationship [11] datasets. The ...

BGP is the de-facto Internet routing protocol for exchanging prefix reachability information between Autonomous Systems (AS). It is a dynamic, distributed, path-vector protocol that enables rich expressions of network policies (typically treated as secrets). In this regime, where complexity is interwoven with information hiding, answering questions such as "what is the expected catchment of the anycast sites of a content provider on the AS-level, if new sites are deployed?", or "how will load-balancing behave if an ISP changes its routing policy for a prefix?", is a hard challenge. In this work, we present a formal model and methodology that takes into account policy-based routing and topological properties of the Internet graph, to predict the routing behavior of networks. We design algorithms that provide new capabilities for informative route inference (e.g., isolating the effect of randomness that is present in prior simulation-based approaches). We analyze the properties of these inference algorithms, and evaluate them using publicly available routing datasets and real-world experiments. The proposed framework can be useful in a number of applications: measurements, traffic engineering, network planning, Internet routing models, etc. As a use case, we study the problem of selecting a set of measurement vantage points to maximize route inference. Our methodology is general and can capture standard valley-free routing, as well as more complex topological and routing setups appearing in practice.Example A: A regional ISP R, whose network spans a region of two major cities, cit A and cit B , has a single upstream tier-1 ISP T A and connects to it at cit A . To avoid overloading its infrastructure in cit A , R decides to connect to another tier-1 ISP T B at cit B . However, after connecting with T B , R observes that 90% of the incoming Internet traffic still enters its network at cit A , therefore the new setup fails to balance R's load among its infrastructure in the two cities. In fact, how to select a transit provider, is a question that lacks a clear answer, and engages operators in active discussions [35]. Example B: A content provider C applies IP anycast (i.e., announces the same IP prefix) [9,13,29,33,49] from three sites. Due to traffic increase, C decides to add one more anycast site. It needs to select where to deploy and how to connect the new site, in order to best split the traffic among its sites. The ongoing research in IP anycasting, e.g., [13,28,49], indicates that this is a problem that is not well-understood yet.While a network can partially determine how other networks route traffic to it through passive (e.g., BGP data [37]) or active (e.g., traceroute, ping) measurements [2,9,13,27,32,49], measurements can provide information only for an existing deployment. However, in many applications (traffic engineering, peering decisions, network resilience, etc.) [30], it is important to know, i.e., predict, how the routing behavior of other networks will change in advance, before a network actually...

In this work, we propose utilizing the rich connectivity between IXPs and ISPs for inter-domain path stitching, supervised by centralized QoS brokers. In this context, we highlight a novel abstraction of the Internet topology, i.e., the inter-IXP multigraph composed of IXPs and paths crossing the domains of their shared member ISPs. This can potentially serve as a dense Internet-wide substrate for provisioning guaranteed end-to-end (e2e) services with high path diversity and global IPv4 address space reach. We thus map the IXP multigraph, evaluate its potential, and introduce a rich algorithmic framework for path stitching on such graph structures.