Measurement Informed Route Selection

2011

Internet routers can obtain distinct routes to the same destination from multiple routing protocols. A mechanism called administrative distance (AD) has evolved to implement route selection in such cases. AD simply imposes a preference ranking among routing protocols. A related mechanism called route redistribution (RR) is used to increase connectivity by injecting routes from one protocol into another. Recent research has shown that the current design of RR with AD mechanisms can give rise to routing and forwarding anomalies that are difficult to debug. We present a new algebraic model that captures not only RR with AD, but also sub-protocol interactions that exist within protocols such as OSPF and IS-IS. Compared with previous work, our model provides a clearer view of the issues and trade-offs involved.

2008

As pointed out by Lakshminarayana et al. [3], there is a tussle between end-users and the ISPs in wanting to control the routes on the Internet. We reexamine the tussle among senders, receivers, and ISPs in light of a few new proposals in routing architectures and propose several directions for future work. End-users or edge networks have limited control over routes. End-users today have little control over how traffic reaches the intended destination. Most end-users residing in edge networks just rely on their upstream provider to deliver outgoing traffic to the destination networks. Even Multihoming provides only limited control in the form of choosing the first hop AS among limited choices for outgoing packets. Such decision is often made based on perceived performance and network cost to select the upstream provider to use for traffic for a given destination. Similarly, for incoming traffic, end-users today do not have much control in terms of the paths traversed by packets. Multi-homed customers can advertise different prefixes across different provider links, use AS path prepending, or advertise specific MED values across links to the same provider in an attempt to perform inbound traffic engineering. To overcome these limitations, overlay routing [1] enables end-hosts to select desired application-layer path traversing through intermediate end-hosts. Such overlay paths, however, cannot avoid the first hop AS as determined the immediate upstream providers. Furthermore, the paths between end-hosts serving as overlay nodes are still controlled by the network. Sender-based control: Given the existing limitations imposed by the networks, end-users have little flexibility in managing both outbound and inbound traffic. There are several proposals to increase route diversity by allowing multi-path BGP [8] and enabling source routing for end-hosts [9]. If end-hosts are exposed multiple paths, they can achieve more flexible control over their paths to arbitrary destination networks for purposes such as load-balancing traffic and gaining better fault-tolerance. Receiver-based control: To perform receiver-based admission control or traffic management, proposals such as capability-based scheme [10] or "off by default" [2] attempt to enable receivers to decide which sender is permitted to send traffic to them. The recent work of FastPass [7] enhances the capability-based approach by imposing traffic rate-limiting along the path from specific senders. To summarize, end-hosts today have limited control over routes for reaching external destinations and for external destinations to reach themselves. The control desired by end-hosts can be generalized as follows. (1) Sender: the endto-end route used to reach a destination. (2) Receiver: the end-to-end route used by a sender host. Note that clearly there can be a conflict between receiver-based and sender-based control; however, some balance can be reached. In terms of detailed control, end-hosts often desire control in the form of permitting certain hosts to send traffic to them, the number of routes the end-host has access to, the ability to bypass a particular network along a path, traffic differentiation among different destinations or senders sharing the same links, the performance characteristics, e.g., fault-tolerance, delay, loss, jitter of the path. Despite potential conflict between sender-based and receiver-based control, receiver should have priority to determine the access control of which senders are allowed to send traffic to it. Ideally, this control should be implemented within the network, close to the sender. Next, we discuss the control over routes from network's perspective.

ACM SIGCOMM Computer Communication Review, 1999

The path taken by a packet traveling across the Internet depends on a large number of factors, including routing protocols and pernetwork routing policies. The impact of these factors on the endto-end performance experienced by users is poorly understood. In this paper, we conduct a measurement-based study comparing the performance seen using the "default" path taken in the Internet with the potential performance available using some alternate path. Our study uses five distinct datasets containing measurements of "path quality", such as round-trip time, loss rate, and bandwidth, taken between pairs of geographically diverse Internet hosts. We construct the set of potential alternate paths by composing these measurements to form new synthetic paths. We find that in 30-80% of the cases, there is an alternate path with significantly superior quality. We argue that the overall result is robust and we explore two hypotheses for explaining it.

IEEE/ACM Transactions on Networking, 1999

Routing policies used in the Internet can be restrictive, limiting communication between source-destination pairs to one path, when often better alternatives exist. To avoid route flapping, recovery mechanisms may be dampened, making adaptation slow. Unstructured overlays have been proposed to mitigate the issues of path and performance failures in the Internet by routing through an indirect-path via overlay peer(s). Choosing alternate-paths in overlay networks is a challenging issue. Ensuring both availability and performance guarantees on alternate paths requires aggressive monitoring of all overlay paths using active probing; this limits scalability. An alternate technique to select an overlay-path is to bias its selection based on physical disjointness criteria to bypass the failure on the primary-path. Recently, several techniques have emerged which can optimize the selection of a disjoint-path without incurring the high costs associated with probing paths. In this paper, we sh...

Proc. of Information Technology: Coding and …

1. Introduction The Internet came into existence in late 70's. Since then the Internet has been growing at an unprecedented rate. Traffic in the Internet continues to double every year [1]. This leads to the need for more bandwidth, better quality and security. Increase in traffic ...

2018 21st Conference on Innovation in Clouds, Internet and Networks and Workshops (ICIN), 2018

Resilience as well as other objectives like censorshipresistance demand the existence of multiple diverse paths between two hosts on the Internet. On Internet level, this requires the use of an overlay approach. In this paper, we study this problem on the basis of a data set obtained from traces between Internet hosts on Planetlab. We study a variety of path selection algorithms, including adapted versions of Suurballe's algorithm. We find that these can outperform the single-hop relay approach that is more commonly proposed for Internet multipath. Our analysis framework imports traceroute data and it generates graph representations of the overlay and the underlay. We can reduce the granularity of information that the algorithms can access. Topology-aware algorithms like Suurballe profit from better information. We also support a variety of evaluation metrics.

ACM Transactions on Modeling and Performance Evaluation of Computing Systems

Interactive and multimedia applications depend on the stability of end-to-end paths for predictable performance and good quality of service. On the other hand, network providers depend on multiple paths to ensure fault tolerance and use load balancing between these paths to enhance the overall network throughput. In this study, we analyze path dynamics for both end-to-end paths and path segments within service providers’ networks using 2 months of measurement data from the RIPE Atlas platform, which collects path traces between a fixed set of source and destination pairs every 15 minutes. We observe that 78% of the end-to-end routes have at least two alternative Autonomous System (AS) paths with some end-to-end routes going through hundreds of different AS paths during the 2 months of analysis. While AS level paths are often prevalent for a day, there are considerable changes in the routing of the trace packets over the ASes for a longer duration of a month or longer. Analyzing end-...

Lecture Notes in Computer Science, 2001

Multipath routing schemes distribute traffic among multiple paths instead of routing all the traffic along a single path. Two key questions that arise in multipath routing are how many paths are needed and how to select these paths. Clearly, the number and the quality of the paths selected dictate the performance of a multipath routing scheme. We address these issues in the context of the proportional routing paradigm where the traffic is proportioned among a few "good" paths instead of routing it all along the "best" path. We propose a hybrid approach that uses both globally exchanged link state metrics-to identify a set of good paths, and locally collected path state metrics-for proportioning traffic among the selected paths. We compare the performance of our approach with that of global optimal proportioning and show that the proposed approach yields near-optimal performance using only a few paths. We also demonstrate that the proposed scheme yields much higher throughput with much smaller overhead compared to other schemes based on link state updates.

Procedia Computer Science, 2018

Network path selection defines the methodology of selecting the best routes and forwarding traffic in a network service provider (NSP). NSPs use routing protocols that are optimized for a single arbitrary metric (i.e., administrative weight), which is commonly calculated according to the link state information, to select network paths. Despite the advantages, link-state protocols lack the ability to select network paths by considering the states of network devices such as the effect of routers for network path selection. Apparently, studying possible techniques for selecting network paths based on the state information of network devices, e.g., routers, has become obligatory. This paper studies the effectiveness of using network device state information for network path selection by presenting the Extended Server Link Router state Routing Protocol (ESLR). ESLR uses the state information of network devices as parameters to calculate a composite route metric; ESLR selects the network paths based on the cumulative propagation delay. By simulating ESLR using an ISP topology, the proposed protocol is examined for the effectiveness of using network device state information for network path selection.