Programming Line-Rate Routers

Channel:
United States Google TechTalks
Subscribers:
348,000
Published on ● Video Link: https://www.youtube.com/watch?v=df2_72wjEdw


Duration: 59:09
3,202 views
32

A Google TechTalk, 10/5/16, Presented by Anirudh Sivaraman
ABSTRACT: The evolution of network routers and switches has been driven primarily by performance. Recently, thanks in part to the emergence of large data centers, the need for better control over network operations, and the desire for new features, programmability of switches has become as important as performance. In response, researchers and practitioners have developed reconfigurable switching chips that are performance-competitive with line-rate fixed-function switching chips. These chips provide some programmability through restricted hardware primitives that can be configured with software directives.

This talk will focus on two abstractions for programming such chips. The first abstraction, packet transactions, lets programmers express packet processing in an imperative language under the illusion that the switch processes exactly one packet at a time. A compiler then translates this sequential programmer view into a pipelined implementation on a switching chip that processes multiple packets concurrently. The second abstraction, a push-in first-out queue, allows programmers to program new scheduling algorithms using a priority queue coupled with a program to compute each packet's priority in the priority queue. For the first time, these two abstractions allow us to program several packet-processing functions at line rate. These packet-processing functions include in-network congestion control, active queue management, data-plane load balancing, network measurement, and packet scheduling.

This talk includes joint work with collaborators at MIT, Barefoot Networks, Cisco Systems, Microsoft Research, Stanford, and University of Washington.

SPEAKER BIO: Anirudh Sivaraman is a PhD student at MIT working in the Computer Science and Artificial Intelligence Laboratory (CSAIL). Before coming to MIT, he was an undergraduate in the Computer Science and Engineering department at IIT Madras and graduated with a BTech in 2010. http://web.mit.edu/anirudh/www/



Other Videos By Google TechTalks


2016-10-20AQC 2016 - An Optimal Stopping Approach for Benchmarking Probabilistic Optimizers
2016-10-20AQC 2016 - Quantum Annealing via Environment-Mediated Quantum Diffusion
2016-10-20AQC 2016 - Inhomogeneous Quasi-adiabatic Driving of Quantum Critical Dynamics
2016-10-20AQC 2016 - Quantum Monte Carlo vs Tunneling vs. Adiabatic Optimization
2016-10-20AQC 2016 - Simulated Quantum Annealing Can Be Exponentially Faster Than Classical
2016-10-20AQC 2016 - Quantum Monte Carlo Simulations and Quantum Annealing
2016-10-20AQC - 2016 Quantum vs. Classical Optimization - A Status Report on the Arms Race
2016-10-20AQC 2016 - What is the Computational Value of Finite Range Tunneling?
2016-10-20AQC 2016 - Opening Remarks: Why We Believe Quantum Annealing Will Succeed
2016-10-14Tracing Android
2016-10-12Programming Line-Rate Routers
2016-09-15Grammatical Framework: Formalizing the Grammars of the World
2016-09-06ZuriHac2016: Parallelizing and Distributing Scientific Software in Haskell
2016-09-06ZuriHac 2016 - Meet Hadrian: A New Build System for GHC
2016-09-06ZuriHac 2016: Generic (and type-level) Programming with Generics-sop
2016-09-06ZuriHac 2016 - Low-level Haskell: An Interactive Tour Through the STG
2016-09-06ZuriHac 2016: Monad Homomorphisms
2016-09-06ZuriHac 2016: Functional Programming at LumiGuide
2016-09-06ZuriHac 2016: Spock - Powerful Elegant Web Applications
2016-05-31Dreamscapes (DeepDream scaled up to Operate on Multi-Hundred Megapixel Images)
2016-05-16Flutter's Rendering Pipeline


Tags:
google techtalk
computer networks
computer routers
network routers