Scaling Internet Routers Using Optics Isaac Keslassy, Shang-Tse Da Chuang, Kyoungsik Yu, David Miller, Mark Horowitz, Olav Solgaard, Nick McKeown Department.

Slides:



Advertisements
Similar presentations
1 EE384Y: Packet Switch Architectures Part II Load-balanced Switch (Borrowed from Isaac Keslassys Defense Talk) Nick McKeown Professor of Electrical Engineering.
Advertisements

1 Maintaining Packet Order in Two-Stage Switches Isaac Keslassy, Nick McKeown Stanford University.
Configuring a Load-Balanced Switch in Hardware Srikanth Arekapudi, Shang-Tse (Da) Chuang, Isaac Keslassy, Nick McKeown Stanford University.
Clean Slate Design for the Internet Designing a Predictable Backbone Network with Valiant Load Balancing NSF 100 x 100 Clean.
Wavelength-Routing Switch Fabric Patrick Chiang, Hossein Kakvand, Milind Kopikare, Uma Krishnamoorthy, Paulina Kuo, Pablo Molinero-Fernández Stanford University.
High-Performance Networking Group Isaac Keslassy, Nick McKeown
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 High Speed Router Design Shivkumar Kalyanaraman Rensselaer Polytechnic Institute
Nick McKeown CS244 Lecture 6 Packet Switches. What you said The very premise of the paper was a bit of an eye- opener for me, for previously I had never.
Frame-Aggregated Concurrent Matching Switch Bill Lin (University of California, San Diego) Isaac Keslassy (Technion, Israel)
A Load-Balanced Switch with an Arbitrary Number of Linecards Isaac Keslassy, Shang-Tse Chuang, Nick McKeown.
Scaling Internet Routers Using Optics UW, October 16 th, 2003 Nick McKeown Joint work with research groups of: David Miller, Mark Horowitz, Olav Solgaard.
May 28th, 2002Nick McKeown 1 Scaling routers: Where do we go from here? HPSR, Kobe, Japan May 28 th, 2002 Nick McKeown Professor of Electrical Engineering.
MEMS and its Applications Optical Routing, an example Shashi Mysore Computer Science UCSB.
Designing Networks with Little or No Buffers or Can Gulliver Survive in Lilliput? Yashar Ganjali High Performance Networking Group Stanford University.
Isaac Keslassy, Shang-Tse (Da) Chuang, Nick McKeown Stanford University The Load-Balanced Router.
A Scalable Switch for Service Guarantees Bill Lin (University of California, San Diego) Isaac Keslassy (Technion, Israel)
Making Parallel Packet Switches Practical Sundar Iyer, Nick McKeown Departments of Electrical Engineering & Computer Science,
1 Circuit Switching in the Core OpenArch April 5 th 2003 Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford University
The Concurrent Matching Switch Architecture Bill Lin (University of California, San Diego) Isaac Keslassy (Technion, Israel)
Scaling Internet Routers Using Optics Producing a 100TB/s Router Ashley Green and Brad Rosen February 16, 2004.
1 Architectural Results in the Optical Router Project Da Chuang, Isaac Keslassy, Nick McKeown High Performance Networking Group
1 OR Project Group II: Packet Buffer Proposal Da Chuang, Isaac Keslassy, Sundar Iyer, Greg Watson, Nick McKeown, Mark Horowitz
Using Load-Balancing To Build High-Performance Routers Isaac Keslassy, Shang-Tse (Da) Chuang, Nick McKeown Stanford University.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion MSM.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion The.
Guaranteed Smooth Scheduling in Packet Switches Isaac Keslassy (Stanford University), Murali Kodialam, T.V. Lakshman, Dimitri Stiliadis (Bell-Labs)
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion Scaling.
A Load-Balanced Switch with an Arbitrary Number of Linecards Isaac Keslassy, Shang-Tse (Da) Chuang, Nick McKeown Stanford University.
Scheduling Proposals Scheduling Group Giulio Galante, Wensheng Hua, Sundar Iyer, Isaac Keslassy, Pablo Molinero, Gireesh Shrimali, Rui Zhang.
The Crosspoint Queued Switch Yossi Kanizo (Technion, Israel) Joint work with Isaac Keslassy (Technion, Israel) and David Hay (Politecnico di Torino, Italy)
1 Internet Routers Stochastics Network Seminar February 22 nd 2002 Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford University.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion Introduction.
Nick McKeown 1 Memory for High Performance Internet Routers Micron February 12 th 2003 Nick McKeown Professor of Electrical Engineering and Computer Science,
1 EE384Y: Packet Switch Architectures Part II Load-balanced Switches Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford University.
Fundamental Complexity of Optical Systems Hadas Kogan, Isaac Keslassy Technion (Israel)
1 Trend in the design and analysis of Internet Routers University of Pennsylvania March 17 th 2003 Nick McKeown Professor of Electrical Engineering and.
CS 268: Computer Networking L-8 Routers. 2 Forwarding and Routers Forwarding IP lookup High-speed router architecture Readings [McK97] A Fast Switched.
1 Achieving 100% throughput Where we are in the course… 1. Switch model 2. Uniform traffic  Technique: Uniform schedule (easy) 3. Non-uniform traffic,
Optimal Load-Balancing Isaac Keslassy (Technion, Israel), Cheng-Shang Chang (National Tsing Hua University, Taiwan), Nick McKeown (Stanford University,
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion Maximal.
August 20 th, A 2.5Tb/s LCS Switch Core Nick McKeown Costas Calamvokis Shang-tse Chuang Accelerating The Broadband Revolution P M C - S I E R R.
15-744: Computer Networking L-8 Routers. 2 Forwarding and Routers Forwarding IP lookup High-speed router architecture Readings [McK97] A Fast Switched.
1 Growth in Router Capacity IPAM, Lake Arrowhead October 2003 Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford University.
Router Design (Nick Feamster) February 11, Today’s Lecture The design of big, fast routers Partridge et al., A 50 Gb/s IP Router Design constraints.
1 IP routers with memory that runs slower than the line rate Nick McKeown Assistant Professor of Electrical Engineering and Computer Science, Stanford.
Using Load-Balancing To Build High-Performance Routers Isaac Keslassy Ph.D. Oral Examination Department of Electrical Engineering Stanford University.
Load Balanced Birkhoff-von Neumann Switches
Nick McKeown CS244 Lecture 7 Valiant Load Balancing.
Optics in Internet Routers Mark Horowitz, Nick McKeown, Olav Solgaard, David Miller Stanford University
How Emerging Optical Technologies will affect the Future Internet NSF Meeting, 5 Dec, 2005 Nick McKeown Stanford University
Advance Computer Networking L-8 Routers Acknowledgments: Lecture slides are from the graduate level Computer Networks course thought by Srinivasan Seshan.
Designing Packet Buffers for Internet Routers Friday, October 23, 2015 Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford.
Winter 2006EE384x1 EE384x: Packet Switch Architectures I Parallel Packet Buffers Nick McKeown Professor of Electrical Engineering and Computer Science,
Applied research laboratory 1 Scaling Internet Routers Using Optics Isaac Keslassy, et al. Proceedings of SIGCOMM Slides:
Nick McKeown1 Building Fast Packet Buffers From Slow Memory CIS Roundtable May 2002 Nick McKeown Professor of Electrical Engineering and Computer Science,
1 Performance Guarantees for Internet Routers ISL Affiliates Meeting April 4 th 2002 Nick McKeown Professor of Electrical Engineering and Computer Science,
1 Router Design Bruce Davie with help from Hari Balakrishnan & Nick McKeown.
Shivkumar Kalyanaraman Rensselaer Polytechnic Institute 1 Challenges in Modern Multi-Tera- bit Class Switch Design.
Opticomm 2001Nick McKeown1 Do Optics Belong in Internet Core Routers? Keynote, Opticomm 2001 Denver, Colorado Nick McKeown Professor of Electrical Engineering.
Challenges in the Next Generation Internet Xin Yuan Department of Computer Science Florida State University
1 A quick tutorial on IP Router design Optics and Routing Seminar October 10 th, 2000 Nick McKeown
1 How scalable is the capacity of (electronic) IP routers? Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford University
A Load Balanced Switch with an Arbitrary Number of Linecards I.Keslassy, S.T.Chuang, N.McKeown ( CSL, Stanford University ) Some slides adapted from authors.
A Load-Balanced Switch with an Arbitrary Number of Linecards Offense Anwis Das.
1 Building big router from lots of little routers Nick McKeown Assistant Professor of Electrical Engineering and Computer Science, Stanford University.
Weren’t routers supposed
The Optical Internet WDM/TDM fiber links Backbone ISP POP, CO Backbone
Parallelism in Network Systems Joint work with Sundar Iyer
CS 740: Advance Computer Networks Hand-out on Router Design
Advance Computer Networking
Presentation transcript:

Scaling Internet Routers Using Optics Isaac Keslassy, Shang-Tse Da Chuang, Kyoungsik Yu, David Miller, Mark Horowitz, Olav Solgaard, Nick McKeown Department of Electrical Engineering Stanford University

Backbone router capacity 1Tb/s 1Gb/s 10Gb/s 100Gb/s Router capacity per rack 2x every 18 months

Backbone router capacity 1Tb/s 1Gb/s 10Gb/s 100Gb/s Router capacity per rack 2x every 18 months Traffic 2x every year

Extrapolating 1Tb/s Router capacity 2x every 18 months Traffic 2x every year 100Tb/s 2015: 16x disparity

Consequence  Unless something changes, operators will need:  16 times as many routers, consuming  16 times as much space,  256 times the power,  Costing 100 times as much.  Actually need more than that…

Stanford 100Tb/s Internet Router Goal: Study scalability  Challenging, but not impossible  Two orders of magnitude faster than deployed routers  We will build components to show feasibility 40Gb/s OpticalSwitch Line termination IP packet processing Packet buffering Line termination IP packet processing Packet buffering Electronic Linecard #1 Electronic Linecard #1 Electronic Linecard #625 Electronic Linecard # Gb/s 160Gb/s Gb/s 100Tb/s = 640 * 160Gb/s

Throughput Guarantees  Operators increasingly demand throughput guarantees:  To maximize use of expensive long-haul links  For predictability and planning  Despite lots of effort and theory, no commercial router today has a throughput guarantee.

Requirements of our router  100Tb/s capacity  100% throughput for all traffic  Must work with any set of linecards present  Use technology available within 3 years  Conform to RFC 1812

What limits router capacity? Approximate power consumption per rack Power density is the limiting factor today

Crossbar Linecards Switch Linecards Trend: Multi-rack routers Reduces power density

Alcatel 7670 RSP Juniper TX8/T640 TX8 Chiaro Avici TSR

Limits to scaling  Overall power is dominated by linecards  Sheer number  Optical WAN components  Per packet processing and buffering.  But power density is dominated by switch fabric

Trend: Multi-rack routers Reduces power density Switch Linecards Limit today ~2.5Tb/s  Electronics  Scheduler scales <2x every 18 months  Opto-electronic conversion

In Out WAN Linecard In WAN Multi-rack routers Out Switch fabric

Question  Instead, can we use an optical fabric at 100Tb/s with 100% throughput?  Conventional answer: No.  Need to reconfigure switch too often  100% throughput requires complex electronic scheduler.

Outline  How to guarantee 100% throughput?  How to eliminate the scheduler?  How to use an optical switch fabric?  How to make it scalable and practical?

In Out R R R R R R Router capacity = NR Switch capacity = N 2 R 100% Throughput ? ? ? ? ? ? ? ? ? R R R R R R R R R R R R R

R In Out R R R R R R/N If traffic is uniform R R

Real traffic is not uniform R In Out R R R R R R/N R R R R R R R R R ?

Out R R R R/N Two-stage load-balancing switch Load-balancing stageSwitching stage In Out R R R R/N R R R 100% throughput for weakly mixing, stochastic traffic. [C.-S. Chang, Valiant]

Out R R R R/N In R R R R/N

Out R R R R/N In R R R R/N

Chang’s load-balanced switch Good properties % throughput for broad class of traffic 1. No scheduler needed  Scalable

Chang’s load-balanced switch Bad properties FOFF: Load-balancing algorithm  Packet sequence maintained  No pathological patterns  100% throughput - always  Delay within bound of ideal  (See paper for details) FOFF: Load-balancing algorithm  Packet sequence maintained  No pathological patterns  100% throughput - always  Delay within bound of ideal  (See paper for details) 1. Packet mis-sequencing 2. Pathological traffic patterns  Throughput 1/N-th of capacity 3. Uses two switch fabrics  Hard to package 4. Doesn’t work with some linecards missing  Impractical

In Out R R R R R R 2R/N Single Mesh Switch One linecard

In R R R Out Backplane R R R Packaging 2R/N R/N

Many fabric options Options Space: Full uniform mesh Time: Round-robin crossbar Wavelength: Static WDM Any permutation network C 1, C 2, …, C N C1C1 C2C2 C3C3 CNCN In Out In Out In Out In Out N channels each at rate 2R/N

In Out In Out In Out In Out Static WDM switching Array Waveguide Router (AWGR) Passive and Almost Zero Power A B C D A, B, C, D A, A, A, A B, B, B, B C, C, C, C D, D, D, D 4 WDM channels, each at rate 2R/N

R WDM 1 N      R Out WDM 1 N 1 N R R      2 R R Linecard dataflow WDM 1 N      RR 3 In

Problems of scale  For N < 64, WDM is a good solution.  We want N = 640.  Need to decompose.

Decomposing the mesh 2R/

Decomposing the mesh 2R/4 2R/ TDM WDM

When N is too large Decompose into groups (or racks) 1, 2, …, G 1 Array Waveguide Router (AWGR) 2L 2R 12L Group/Rack 1 Group/Rack G 1 G 1, 2, …, G

When a linecard is missing  Each linecard spreads its data equally over every other linecard.  Problem: If one is missing, or failed, then the spreading no longer works.

When a linecard fails In Out R R R 2R/3 2R/3 + 2R/3 = 4R/3 In R R R 2R/3 + 2R/6 2R/3 + 2R/6 + 2R/3 + 2R/6 = 2R 2R/3 + 2R/6 Solution: 1.Move light beams  Replace AWGR with MEMS switch.  Reconfigure when linecard added, removed or fails. 2.Finer channel granularity  Multiple paths.

Solution Use transparent MEMS switches 12L 2R 12L Group/Rack 1 Group/Rack G=40 MEMS Switch 1 G MEMS Switch 1 G MEMS Switch 1 G Theorems: 1. Require L+G - 1 MEMS switches 2. Polynomial time reconfiguration algorithm MEMS switches reconfigured only when linecard added, removed or fails.

Challenges R WDM 1 G     G R Out WDM 1 G Pkt Switch 1 G R R     G 2 R=160Gb/s R WDM 1 G     G Address Lookup 11 RR 3 In How to build a 250ms 160Gb/s buffer? Low-cost, low-power optoelectronic conversion?

What we are building Buffer Manager 90nm ASIC Buffer Manager 90nm ASIC 250ms DRAM 160Gb/s 320Gb/s Chip #1: 160Gb/s Packet Buffer CMOS ASIC 16 x 10Gb/s To LinecardsTo Optical Fabric Chip #2: 16 x 55 Opto-electronic crossbar 55 x 10Gb/s Optical source Optical Detector Optical Modulator

100Tb/s Load-Balanced Router L = Gb/s linecards Linecard Rack G = 40 L = Gb/s linecards Linecard Rack 1 L = Gb/s linecards x 40 MEMS Switch Rack < 100W