12:45PM, Thursday, November 17th 2005.
Gates 104

Buffer Sizing in All-Optical Packet Switches

Yashar Ganjali
Stanford University

About the talk:
Internet routers require buffers to hold packets during times of congestion.  The buffers need to be fast, and so ideally they should be small enough to use fast memory technologies such as SRAM or all-optical buffering.  Unfortunately, a widely used rule-of-thumb says we need a bandwidth-delay product of buffering at each router so as not to lose link utilization. This can be prohibitively large. In a recent paper, Appenzeller et al. challenged this rule-of-thumb and showed that for a backbone network, the buffer size can be divided by \sqrt(N) without sacrificing throughput, where N is the number of flows sharing the bottleneck. In this work, we first provide some experimental validation (using data gathered from the Level3 Communications backbone) for the reduced buffer size result. Then, we explore how buffers in the backbone can be significantly reduced even more, to as little as a few dozen packets, if we are willing to sacrifice a small amount of link capacity. We argue that if the TCP sources are not overly bursty, then fewer than twenty packet buffers are sufficient for high throughput. Specifically, we argue that O(\log W) buffers are sufficient, where W is the window size of each flow. We support our claim with analysis and a variety of simulations. The change we need to make to TCP is minimal: each sender just needs to pace packet injections from its window. Moreover, there is some evidence that such small buffers are sufficient even if we don't modify the TCP sources, so long as the access network is much slower than the backbone, which is true today and likely to remain true in the future.  

We conclude that buffers can be made small enough for all-optical routers with small integrated optical buffers. For details, see the SIGCOMM 2004 paper and the ACM/Sigcomm CCR paper.

About the speaker:
Yashar Ganjali is with the High Performance Networking Group at Stanford University, where he is working toward his Ph.D. degree. He received a B.Sc. in Computer Engineering from Sharif University of Technology, Tehran, Iran, in 1999, and an M.Sc. in Computer Science from the University of Waterloo, Waterloo, Canada, in 2001. He is currently involved with the buffer sizing project: its goal is to determine the impact of reducing the buffer size in core routers from millions of packets to just tens of packets, and thus exploring the possibility of building all-optical networks. His other research interests include analysis and design of high performance switches, scheduling algorithms, congestion control, routing protocols, and network optimization.