Summary: Best­Path vs. Multi­Path Overlay Routing
David G. Andersen, Alex C. Snoeren y , and Hari Balakrishnan
MIT Laboratory for Computer Science
fdga,harig@lcs.mit.edu
y University of California, San Diego
snoeren@cs.ucsd.edu
Abstract
Time­varying congestion on Internet paths and failures due to soft­
ware, hardware, and configuration errors often disrupt packet deliv­
ery on the Internet. Many aproaches to avoiding these problems use
multiple paths between two network locations. These approaches
rely on a path­independence assumption in order to work well; i.e.,
they work best when the problems on different paths between two
locations are uncorrelated in time.
This paper examines the extent to which this assumption holds
on the Internet by analyzing 14 days of data collected from 30
nodes in the RON testbed. We examine two problems that mani­
fest themselves---congestion­triggered loss and path failures---and
find that the chances of losing two packets between the same hosts
is nearly as high when those packets are sent through an interme­

Source: Andersen, Dave - School of Computer Science, Carnegie Mellon University
Massachusetts Institute of Technology (MIT), Computer Science and Artificial Intelligence Laboratory, Networks and Mobile Systems Group

Collections: Computer Technologies and Information Sciences