Overlay Networks: An Akamai Perspective

The Internet is transforming every aspect of communication in human society by enabling a wide range of applications for business, commerce, entertainment, news, and social interaction. Modern and future distributed applications require high reliability, performance, security, and scalability, and yet need to be developed rapidly and sustainably at low operating costs. For instance, major e-commerce sites require at least “four nines” (99.99%) of reliability, allowing no more than a few minutes of downtime per month. As another example, the future migration of high-quality television to the Internet would require massive scalability to flawlessly transport tens of petabits per second of data to global audiences around the world. However, the Internet was never architected to provide the stringent requirements of such modern and futuristic distributed applications. It was created as a heterogeneous network of networks, and its design enables various entities to interact with each other in a “best effort” fashion. Guarantees on high performance, availability, scalability and security are not inherently provided on the Internet in accordance with its best effort design principle. Today’s Internet is a vast patchwork of more than 13,000 autonomous networks that often compete for business. Failures and performance degradation in transporting information across this patchwork are routine occurrences. So, how would we bridge the gap between what modern Internet-based services need and what the Internet actually provides? A complete clean-

[1]  Hari Balakrishnan,et al.  Resilient overlay networks , 2001, SOSP.

[2]  Ben Y. Zhao,et al.  Tapestry: a resilient global-scale overlay for service deployment , 2004, IEEE Journal on Selected Areas in Communications.

[3]  Ramesh K. Sitaraman,et al.  A transport layer for live streaming in a content delivery network , 2004, Proceedings of the IEEE.

[4]  Bruce M. Maggs,et al.  Globally Distributed Content Delivery , 2002, IEEE Internet Comput..

[5]  Scott Shenker,et al.  Making gnutella-like P2P systems scalable , 2003, SIGCOMM '03.

[6]  Arjang A. Assad,et al.  Multicommodity network flows - A survey , 1978, Networks.

[7]  Hans Eriksson,et al.  MBONE: the multicast backbone , 1994, CACM.

[8]  Micah Adler,et al.  Algorithms for optimizing the bandwidth cost of content delivery , 2011, Comput. Networks.

[9]  Bruce M. Maggs,et al.  On the fault tolerance of some popular bounded-degree networks , 1992, Proceedings., 33rd Annual Symposium on Foundations of Computer Science.

[10]  David R. Karger,et al.  Chord: A scalable peer-to-peer lookup service for internet applications , 2001, SIGCOMM '01.

[11]  Mark Handley,et al.  A scalable content-addressable network , 2001, SIGCOMM '01.

[12]  Arnold L. Rosenberg,et al.  Work-preserving emulations of fixed-connection networks , 1989, STOC '89.

[13]  Antony I. T. Rowstron,et al.  Pastry: Scalable, Decentralized Object Location, and Routing for Large-Scale Peer-to-Peer Systems , 2001, Middleware.

[14]  Bruce M. Maggs,et al.  Designing overlay multicast networks for streaming , 2003, SPAA '03.

[15]  Bruce M. Maggs,et al.  Algorithms for Constructing Overlay Networks For Live Streaming , 2011, ArXiv.

[16]  Peter Druschel,et al.  Pastry: Scalable, distributed object location and routing for large-scale peer-to- , 2001 .

[17]  Ramesh K. Sitaraman,et al.  Communication and fault tolerance in parallel computers , 1993 .

[18]  Anees Shaikh,et al.  A measurement-based analysis of multihoming , 2003, SIGCOMM '03.

[19]  Amin Vahdat,et al.  Detour: a Case for Informed Internet Routing and Transport , 2007 .

[20]  Kimberly C. Claffy,et al.  Measuring the deployment of IPv6: topology, routing and performance , 2012, IMC '12.

[21]  Ramesh K. Sitaraman,et al.  The Akamai network: a platform for high-performance internet applications , 2010, OPSR.