Explaining BGP Slow Table Transfers

Abstract

Although there have been a plethora of studies on TCP performance in supporting of various applications, relatively little is known about the interaction between TCP and BGP, which is a specific application running on top of TCP. This paper investigates BGP's slow route propagation by analyzing packet traces collected from a large ISP and Route Views Oregon collector. In particular we focus on the prolonged periods of BGP routing table transfers and examine in detail the interplay between TCP and BGP. In addition to the problems reported in previous literature, this study reveals a number of new TCP transport problems, that collectively induce significant delays. Furthermore, we develop a tool, named T-DAT, that can be deployed together with BGP data collectors to infer various factors behind the observed delay, including BGP's sending and receiving behavior, TCP's parameter settings, TCP's flow and congestion control, and network path limitation. Identifying these delay contributing factors makes an important step for ISPs and router vendors to diagnose and improve the BGP performance.