A measurement study on the impact of routing events on end-to-end internet path performance

Abstract

Extensive measurement studies have shown that end-to-end Internet path performance degradation is correlated with routing dynamics. However, the root cause of the correlation between routing dynamics and such performance degradation is poorly understood. In particular, how do routing changes result in degraded end-toend path performance in the first place? How do factors such as topological properties, routing policies, and iBGP configurations affect the extent to which such routing events can cause performance degradation? Answers to these questions are critical for improving network performance.In this paper, we conduct extensive measurement that involves both controlled routing updates through two tier-1 ISPs and active probes of a diverse set of end-to-end paths on the Internet. We find that routing changes contribute to end-to-end packet loss significantly. Specifically, we study failover events in which a link failure leads to a routing change and recovery events in which a link repair causes a routing change. In both cases, it is possible to experience data plane performance degradation in terms of increased long loss burst as well as forwarding loops. Furthermore, we find that common routing policies and iBGP configurations of ISPs can directly affect the end-to-end path performance during routing changes. Our work provides new insights into potential measures that network operators can undertake to enhance network performance.

An investigation of the Internet’s IP-layer connectivity

“…While effective on reducing the number of updates triggered by each routing event, MRAI timer can also defer routing convergence, and if implemented improperly, it can even lead to long-lived routing black holes. Previous study has revealed that various inter-domain routing issues including routing policies, iBGP configurations, MRAI timer values, and failure locations, can all have significant impact on the Internet's routing failures [32]. In regard to our experiment, a majority of the IP-layer unreachability caused by inter-domain routing dynamics was most likely classified into category C2, and a small portion into C4.…”

“…However, due to the diverse distribution of PlanetLab nodes, we believe the above statistics are still representative. Moreover, there have been many other measurement studies correlating the routing instability and E2E performance [32,27,[33][34][35]. All these indicate that routing dynamics can lead to IP-layer unreachability lasting for a wide variation of time.…”

Layered multicast for reliable event notification over large-scale networks

“…• WANs exhibit not-negligible bursty loss patterns [52] , i.e. , a packet has a considerable probability to be lost during the delivery and more than one packet can be dropped consecutively.…”

Reducing burst packet loss through route-free forwarding

“…Labovitz, et al report that on average a route change generates 30% packet loss for as long as two minutes [1] . Wang, et al show that burst packet loss that last up to 20 seconds after a single route failure and up to 8 seconds after a route recovery event [2] . In today's high speed networks, even a short time can cause huge packets loss.…”

“…The distribution nature of Internet routing decisions and the rate-limiting timer Minimum Route Advertisement Interval (MRAI) [4] of BGP introduce unavoidable latency in reaction to network changes. The asynchronous route computation results in temporarily routing table inconsistencies which may cause short-term routing blackholes and loops [2,5,6] . We refer to routing blackholes and routing loops as routing anomaly which contribute to the burst of packet loss.…”