A Comparison of Overlay Routing and Multihoming Route Control
by Aditya Akella, Jeff Pang, Anees Shaikh, Srinivasan Seshan, Bruce Maggs
url show details
Details
| type: | misc | | publisher: | ACM | | year: | 2004 | | url: | http://pages.cs.wisc.edu/~akella/CS740/F08/740-Papers/APM+04.pdf | | pages: | 93--106 | | abstract: | The limitations of BGP routing in the Internet are often blamed for
poor end-to-end performance and prolonged connectivity interruptions.
Recent work advocates using overlays to effectively bypass
BGP’s path selection in order to improve performance and fault
tolerance. In this paper, we explore the possibility that intelligent
control of BGP routes, coupled with ISP multihoming, can provide
competitive end-to-end performance and reliability. Using extensive
measurements of paths between nodes in a large content distribution
network, we compare the relative benefits of overlay routing
and multihoming route control in terms of round-trip latency, TCP
connection throughput, and path availability. We observe that the
performance achieved by route control together with multihoming
to three ISPs (3-multihoming), is within 5-15% of overlay routing
employed in conjunction 3-multihoming, in terms of both endto-
end RTT and throughput. We also show that while multihoming
cannot offer the nearly perfect resilience of overlays, it can
eliminate almost all failures experienced by a singly-homed endnetwork.
Our results demonstrate that, by leveraging the capability
of multihoming route control, it is not necessary to circumvent
BGP routing to extract good wide-area performance and availability
from the existing routing system. | | booktitle: | Proceedings of the {ACM} {SIGCOMM} 2004 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication, August 30 - September 3, 2004, Portland, Oregon, {USA} |
|
|
You need to log in to add tags and post comments.
The differences in overlay routing and BGP routing can be attributed to differences in their route availability and route selection schemes. Overlay routing is attractive due to its property of allowing end points to specify a set of intermediate hops. This gives end hosts more control over which paths their packets take. This is an advantage over the policy-driven routing done by BGP where end hosts are constrained to the number of ISP connections it has. I really enjoyed the layout of section 5. Using varying values of k in reference to multihoming and overlays really illuminated the tradeoffs and differences between the two techniques. A key point is that when more ISP’s are used in conjunction with overlay routing, overlay routing selects a high fraction of direct BGP paths. These marginal benefits over multihoming alone emphasizes that it is not totally necessary to go around BGP routing to achieve good end-to-end performance.