Anycast Services Research Articles

RMSA algorithms resilient to multiple node failures in dynamic EONs

In Elastic Optical Networks (EONs), the way different service demands are supported in the network is ruled by the Routing, Modulation and Spectrum Assignment (RMSA) algorithm, which decides how the spectrum resources of the optical network are assigned to each service demand. In a dynamic EON, demand requests arrive randomly one at a time and the accepted demands last in the network for a random time duration. So, one important goal of the RMSA algorithm is the efficient use of the spectrum resources to maximize the acceptance probability of future demand requests. On the other hand, multiple failure events are becoming a concern to network operators as such events are becoming more frequent in time. In this work, we consider the case of multiple node failure events caused by malicious attacks against network nodes. In order to obtain RMSA algorithms resilient to such events, a path disaster availability metric was recently proposed which takes into account the probability of each path not being disrupted by an attack. This metric was proposed in the offline variant of the RMSA problem where all demands are assumed to be known at the beginning. Here, we exploit the use of the path disaster availability metric in the RMSA of dynamic EONs. In particular, we propose RMSA algorithms combining the path disaster availability metric with spectrum usage metrics in a dynamic way based on the network load level. The aim is that the efficient use of the resources is relaxed for improved resilience to multiple node failures when the EON is lightly loaded, while it becomes the most important goal when the EON becomes heavily loaded. We present simulation results considering a mix of unicast and anycast services in 3 well-known topologies. The results show that the RMSA algorithms combining the path disaster availability metric with spectrum usage metrics are the best trade-off between spectrum usage efficiency and resilience to multiple node failures.

Does Anycast Hang Up on You (UDP and TCP)?

Anycast-based services today are widely used commercially, with several major providers serving thousands of important websites. However, to our knowledge, there has been only limited study of how often anycast fails because routing changes interrupt connections between users and their current anycast site. While the commercial success of anycast CDNs means anycast usually works well, do some users end up shut out of anycast? In this paper we examine data from more than 9000 geographically distributed vantage points (VPs) to 11 anycast services to evaluate this question. Our contribution is the analysis of this data to provide the first quantification of this problem, and to explore where and why it occurs. We see that about 1% of VPs are anycast unstable, reaching a different anycast site frequently (sometimes every query). Flips back and forth between two sites in 10 seconds are observed in selected experiments for given service and VPs. Moreover, we show that anycast instability is persistent for some VPs — a few VPs never see a stable connections to certain anycast services during a week or even longer. The vast majority of VPs only saw unstable routing towards one or two services instead of instability with all services, suggesting the cause of the instability lies somewhere in the path to the anycast sites. Finally, we point out that for highly-unstable VPs, their probability to hit a given site is constant, which means the flipping are happening at a fine granularity — per packet level, suggesting load balancing might be the cause to anycast routing flipping. Our findings confirm the common wisdom that anycast almost always works well, but provide evidence that a small number of locations in the Internet where specific anycast services are never stable.

Design and Implementation of an Anycast Services Discovery in Mobile Ad Hoc Networks

Anycasting is a network service that selects the best one of service providers in an anycast group as a destination. While anycasting offers better service flexibility in mobile ad hoc networks (MANETs), it also incurs new problems. In MANETs, every node can move arbitrarily, and the routes from mobile nodes to their service providers would vary. Therefore, anycast service discovery in MANETs usually relies on network-layer message broadcasting, which leads to large traffic overhead for the scarce bandwidth of MANETs. In this work, we present a traffic-control scheme for anycast service discovery in MANETs. Our scheme can reduce the volume of query messages and the reply messages. In addition to basic anycasting, our scheme also supports k-anycast service that requests for k anycast service providers in each service instance. With k-anycast service, the fault tolerance and service flexibility of our scheme can be improved. Experimental results demonstrate that our scheme is efficient and feasible for MANETs.

Design and Implementation of Anycast Services in Ad Hoc Networks Connected to IPv6 Networks

The paper proposes a communication model of implementing an Anycast service in an Ad Hoc network which is connected to IPv6 networks where IPv6 nodes can obtain the Anycast service provided by the Ad hoc network. In this model when an Anycast mobile member in an Ad hoc network moves it can keep the existing communications with its corresponding nodes to continue providing the Anycast services with good quality of service to IPv6 nodes. This model creates a new kind of IPv6 address auto-configuration scheme which does not need the address duplication detection. This paper deeply discusses and analyzes the model and the experimental data prove its validity and efficiency.

SCTP for robust and flexible IP anycast services

IP anycast is a powerful network layer mechanism that can be used for transparent communications between clients and a distributed service infrastructure. Unfortunately, large-scale deployment of IP anycast would cause a number of severe problems, including excessive routing table growth and potential routing instability. In order to solve these problems, a number of overlay network architectures have been proposed over the last years. In this paper we show that the robustness of anycast services provided via such anycast architectures can be significantly improved by using SCTP transport layer facilities. More specifically, the proposed approach adds the following important features to existing anycast overlays: Robustness to anycast overlay node failure or network reconfiguration, seamless anycast service delivery to mobile clients, and true stateful anycast communications over an entirely stateless infrastructure. Furthermore, we argue that the number of overlay nodes can be drastically reduced in comparison with the earlier architectures, without degrading service quality or increasing the end-to-end path stretch.

Design and implementation of Anycast communication model in IPv6

AbstractThe existing designs for providing Anycast services are either to confine each Anycast group to a preconfigured topological region or to distribute members of Anycast groups over global regions. The former brings an Anycast scalability problem and the latter causes the routing tables to grow proportionally to the number of all global Anycast groups in the entire Internet. Therefore, both of the above designs restrict and hinder the application and development of Anycast services. A new kind of Anycast communication model is proposed in this paper which solves some existing problems, such as scalability and communication errors between clients and servers. In this paper, the Anycast communication model is analyzed in depth and discussed, and the experimental data of this Anycast communication model demonstrate its feasibility and validity.

ASTAS: Architecture for scalable and transparent anycast services

Native information provider (IP) anycast suffers from routing scalability issues and the lack of stateful communication support. For this reason, we propose architecture for scalable and transparent anycast services (ASTAS), a proxy-based architecture that provides support for stateful anycast communications, while retaining the transparency offered by native anycast. Dynamic re- source assignment for each initiated session guarantees that a connection is established with the most suitable target server, based on network and server conditions. Traffic engineering in the overlay can be realized in an effective way due to the dissemination of aggregated state information in the anycast overlay. To minimize the total deployment cost for ASTAS architectures, we propose optimized proxy placement and path finding heuristics based on look-ahead information gathered in network nodes. Contrary to a regular integer linear program (ILP) formulation, these heuristics allow to optimize proxy placement in large networks. A use case on a European reference network illustrates that lower proxy costs enable proxy deployment closer to the end-users, resulting in a reduced network load.

A scheme for solving Anycast scalability in IPv6

AbstractThe existing designs for providing Anycast services are either to confine Anycast groups to a preconfigured topological region or to distribute Anycast groups globally across the whole Internet. The latter causes routing tables to grow proportionally to the number of global Anycast groups in the entire Internet and both of the above designs restrict and hinder the application and development of Anycast services. A new kind of Anycast communication scheme is proposed in this paper. This scheme adopts a novel Anycast address structure which can achieve a dynamic Anycast group while allowing Anycast members to freely leave and join the Anycast group without geographical restriction and it effectively solves the expanding explosion of the Anycast routing table. In addition, this scheme can evenly disperse Anycast request messages from clients across the Anycast servers of one Anycast group, thus achieving load balance. This paper analyzes the communication scheme in depth and discusses its feasibility and validity. The experimental data in IPv6 simulation demonstrate that the TRT (Total Response Time) of one Anycast service (e.g., file downloading) acquired through this communication scheme is shorter by 15% than that through the existing Anycast communication scheme. Copyright © 2007 John Wiley & Sons, Ltd.

Analysis and discussion of Anycast scalability in IPv6

AbstractThe existing designs for providing Anycast services are either to confine Anycast groups to a preconfigured topological region or to globally distribute Anycast groups across the whole Internet. The latter causes the routing tables to grow proportionally to the number of global Anycast groups in the entire Internet and both of the above designs restrict and hinder the application and development of Anycast services. In this paper a new kind of Anycast communication model is proposed on the basis of Anycast tree technology. Since this model adopts a dynamic Anycast group, allowing Anycast members to freely leave and join the Anycast group, it radically solves the existing scalability problem. In addition, this model accomplishes the distributed maintenance and transaction of Anycast service requests and information on the Anycast tree; thus it performs load balance. This paper analyzes and discusses the feasibility and validity of the communication model; the experimental data in IPv6 simulation demonstrate that the total response time of one Anycast service (e.g., file downloading) acquired through this communication model is shorter by 15% than that through the current communication model. Copyright © 2007 John Wiley & Sons, Ltd.

A routing protocol for anycast messages

An anycast packet is one that should be delivered to one member in a group of designated recipients. Using anycast services may considerably simplify some applications. Little work has been done on routing anycast packets. In this paper, we propose and analyze a routing protocol for anycast message. It is composed of two subprotocols: the routing table establishment subprotocol and the packet forwarding subprotocol. In the routing table establishment subprotocol, we propose four methods (SSP, MIN-D, SET, and GET) for enforcing an order among routers for the purpose of loop prevention. These methods differ from each other on information used to maintain orders, the impact on QoS, and the compatibility to the existing routing protocols. In the packet forwarding subprotocol, we propose a Weighted-Random Selection (WRS) approach for multiple path selection in order to balance network traffic. In particular, the fixed and adaptive methods are proposed to determine the weights. Both of them explicitly take into account the characteristics of distribution of anycast recipient group while the adaptive method uses the dynamic information of the anycast traffic as well. Correctness property of the protocol is formally proven. Extensive simulation is performed to evaluate our newly designed protocol. Performance data shows that the loop-prevention methods and the WRS approaches have great impact on the performance in terms of average end-to-end packet delay. In particular, the protocol using the SET or CBT loop-prevention methods and the adaptive WRS approach performs very close to a dynamic optimal routing protocol in most cases.