Maximum Size Matchings & Input Queued Switches Sundar Iyer, Nick McKeown High Performance Networking Group, Stanford University,

Slides:



Advertisements
Similar presentations
1 Maintaining Packet Order in Two-Stage Switches Isaac Keslassy, Nick McKeown Stanford University.
Advertisements

1 Scheduling Crossbar Switches Who do we chose to traverse the switch in the next time slot? N N 11.
Intelligent Packet Dropping for Optimal Energy-Delay Tradeoffs for Wireless Michael J. Neely University of Southern California
High-Performance Networking Group Isaac Keslassy, Nick McKeown
Submitters: Erez Rokah Erez Goldshide Supervisor: Yossi Kanizo.
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)
Routers with a Single Stage of Buffering Sundar Iyer, Rui Zhang, Nick McKeown High Performance Networking Group, Stanford University,
Towards Simple, High-performance Input-Queued Switch Schedulers Devavrat Shah Stanford University Berkeley, Dec 5 Joint work with Paolo Giaccone and Balaji.
Isaac Keslassy, Shang-Tse (Da) Chuang, Nick McKeown Stanford University The Load-Balanced Router.
Algorithm Orals Algorithm Qualifying Examination Orals Achieving 100% Throughput in IQ/CIOQ Switches using Maximum Size and Maximal Matching Algorithms.
Making Parallel Packet Switches Practical Sundar Iyer, Nick McKeown Departments of Electrical Engineering & Computer Science,
Fast Matching Algorithms for Repetitive Optimization Sanjay Shakkottai, UT Austin Joint work with Supratim Deb (Bell Labs) and Devavrat Shah (MIT)
1 Input Queued Switches: Cell Switching vs. Packet Switching Abtin Keshavarzian Joint work with Yashar Ganjali, Devavrat Shah Stanford University.
Analysis of a Packet Switch with Memories Running Slower than the Line Rate Sundar Iyer, Amr Awadallah, Nick McKeown Departments.
1 Architectural Results in the Optical Router Project Da Chuang, Isaac Keslassy, Nick McKeown High Performance Networking Group
1 ENTS689L: Packet Processing and Switching Buffer-less Switch Fabric Architectures Buffer-less Switch Fabric Architectures Vahid Tabatabaee Fall 2006.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion MSM.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion The.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion Scaling.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion Statistical.
Scheduling for maximizing throughput EECS, UC Berkeley Presented by Antonis Dimakis
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.
Lecture 11. Matching A set of edges which do not share a vertex is a matching. Application: Wireless Networks may consist of nodes with single radios,
1 EE384Y: Packet Switch Architectures Part II Load-balanced Switches Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford University.
1 Trend in the design and analysis of Internet Routers University of Pennsylvania March 17 th 2003 Nick McKeown Professor of Electrical Engineering and.
1 Achieving 100% throughput Where we are in the course… 1. Switch model 2. Uniform traffic  Technique: Uniform schedule (easy) 3. Non-uniform traffic,
1 Netcomm 2005 Communication Networks Recitation 5.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion Maximal.
Surprise Quiz EE384Z: McKeown, Prabhakar ”Your Worst Nightmares in Packet Switching Architectures”, 3 units [Total time = 15 mins, Marks: 15, Credit is.
048866: Packet Switch Architectures Dr. Isaac Keslassy Electrical Engineering, Technion Scheduling.
Distributed Scheduling Algorithms for Switching Systems Shunyuan Ye, Yanming Shen, Shivendra Panwar
Pipelined Two Step Iterative Matching Algorithms for CIOQ Crossbar Switches Deng Pan and Yuanyuan Yang State University of New York, Stony Brook.
Localized Asynchronous Packet Scheduling for Buffered Crossbar Switches Deng Pan and Yuanyuan Yang State University of New York Stony Brook.
1 IP routers with memory that runs slower than the line rate Nick McKeown Assistant Professor of Electrical Engineering and Computer Science, Stanford.
Load Balanced Birkhoff-von Neumann Switches
Belgrade University Aleksandra Smiljanić: High-Capacity Switching High-Capacity Packet Switches.
Belgrade University Aleksandra Smiljanić: High-Capacity Switching Switches with Input Buffers (Cisco)
High Speed Stable Packet Switches Shivendra S. Panwar Joint work with: Yihan Li, Yanming Shen and H. Jonathan Chao New York State Center for Advanced Technology.
Enabling Class of Service for CIOQ Switches with Maximal Weighted Algorithms Thursday, October 08, 2015 Feng Wang Siu Hong Yuen.
Summary of switching theory Balaji Prabhakar Stanford University.
Routers. These high-end, carrier-grade 7600 models process up to 30 million packets per second (pps).
Packet Forwarding. A router has several input/output lines. From an input line, it receives a packet. It will check the header of the packet to determine.
Abtin Keshavarzian Yashar Ganjali Department of Electrical Engineering Stanford University June 5, 2002 Cell Switching vs. Packet Switching EE384Y: Packet.
1 Performance Guarantees for Internet Routers ISL Affiliates Meeting April 4 th 2002 Nick McKeown Professor of Electrical Engineering and Computer Science,
Stress Resistant Scheduling Algorithms for CIOQ Switches Prashanth Pappu Applied Research Laboratory Washington University in St Louis “Stress Resistant.
Winter 2006EE384x1 EE384x: Packet Switch Architectures I a) Delay Guarantees with Parallel Shared Memory b) Summary of Deterministic Analysis Nick McKeown.
Belgrade University Aleksandra Smiljanić: High-Capacity Switching Switches with Input Buffers (Cisco)
Order Optimal Delay for Opportunistic Scheduling In Multi-User Wireless Uplinks and Downlinks Michael J. Neely University of Southern California
Techniques for Fast Packet Buffers Sundar Iyer, Ramana Rao, Nick McKeown (sundaes,ramana, Departments of Electrical Engineering & Computer.
Buffered Crossbars With Performance Guarantees Shang-Tse (Da) Chuang Cisco Systems EE384Y Thursday, April 27, 2006.
SNRC Meeting June 7 th, Crossbar Switch Scheduling Nick McKeown Professor of Electrical Engineering and Computer Science, Stanford University
Improving Matching algorithms for IQ switches Abhishek Das John J Kim.
Topics in Internet Research: Project Scope Mehreen Alam
Reduced Rate Switching in Optical Routers using Prediction Ritesh K. Madan, Yang Jiao EE384Y Course Project.
Throughput of Internally Buffered Crossbar Switch Saturday, February 20, 2016 Mingjie Lin
Input buffered switches (1)
1 Chapter 7 Network Flow Slides by Kevin Wayne. Copyright © 2005 Pearson-Addison Wesley. All rights reserved.
Analysis of Maximum Size Matching scheduling algorithm (MSM) in input- queued switches under uniform traffic Neda Beheshti and Mohsen Bayati {nbehesht,
scheduling for local-area networks”
Balaji Prabhakar Departments of EE and CS Stanford University
Packet Forwarding.
Chapter 7 Network Flow Slides by Kevin Wayne. Copyright © 2005 Pearson-Addison Wesley. All rights reserved.
Stability Analysis of MNCM Class of Algorithms and two more problems !
Balaji Prabhakar Departments of EE and CS Stanford University
Long Gong, Paul Tune, Liang Liu, Sen, Jun (Jim) Xu
Write about the funding Sundar Iyer, Amr Awadallah, Nick McKeown
EE384Y: Packet Switch Architectures II
Presentation transcript:

Maximum Size Matchings & Input Queued Switches Sundar Iyer, Nick McKeown High Performance Networking Group, Stanford University, Allerton 2002 Wednesday, Oct 2 nd 2002

2 Definition - 100% Throughput A switch gives 100% throughput if the expected size of the queues is finite for any admissible (no input or output is oversubscribed) load.

3 A Characteristic Switch N=4 1 1 R R An input queued switch with a crossbar switching fabric Crossbar R R 1 N=4 1 VOQs

4 Maximum Size Matching  Maximum Size Matching (MSM)  Choose a matching which maximizes the size  Contrary to intuition, MSM does not give 100% throughput Ref: [McKeown, Anantharam, Walrand ], “Achieving 100% Throughput in an Input-Queued Switch“, IEEE Infocom '96.

5 Contents 1. Background & Motivation 2. Non Pre-emptive Scheduling 3. Achieving 100% throughput with CMSM Bernoulli i.i.d. uniform traffic Bernoulli i.i.d. non-uniform traffic 4. Achieving 100% throughput with MSM Bernoulli i.i.d. uniform traffic 5. Conclusion

6 An Example MSM does not give 100% throughput N=2 1 1 R R Crossbar R R  11 =0.49  12 =0.50  21 =0.50  22 =0.00 Ref: [Keslassy, Zhang, McKeown ], “MSM is unstable for any input queued switch”, In Preparation. VOQs

7 Motivation “To understand the conditions under which the class of MSMs give 100% throughput”

8 Questions  Do all MSMs not achieve 100% throughput?  Is there a sub class of MSMs which achieve 100% throughput?  Do all MSMs achieve 100% throughput under uniform load?

9 Contents 1. Background & Motivation 2. Non Pre-emptive Scheduling 3. Achieving 100% throughput with CMSM Bernoulli i.i.d. uniform traffic Bernoulli i.i.d. non-uniform traffic 4. Achieving 100% throughput with MSM Bernoulli i.i.d. uniform traffic 5. Conclusion

10 Non Pre-emptive Scheduling … 1 Batch Scheduling  Main Idea  Scheduling cells in batches increases the choice for the matching and hence increases throughput  Allow the batch size to grow Ref: [Dolev, Kesselman ], “Bounded latency scheduling scheme for ATM cells", Computer Networks, vol. 32(3) pp , 2000.

11 Non Pre-emptive Scheduling … 2 Batch Scheduling N N 1 1 R R Priority-2 Crossbar R R 1 N 1 N Priority-1 Batch- (k+1) Batch- (k)

12 Non Pre-emptive Scheduling … 2 Batch Scheduling N N 1 1 R R Priority-2 R R 1 N 1 N Priority-1 Crossbar Batch- (k+1) Batch- (k)

13 Degree of a Batch Batch Request Graph  Degree ( d v,k ):  The number of cells departing from (destined to) a vertex in batch k.  Maximum Degree ( D k )  The maximum degree amongst all inputs/outputs in batch k.

Batch Request Graph with D k = Maximum Size Matching Why may MSM not give 100% throughput?

15 Critical Maximum Size Matching A sub-class of MSM Batch Request Graph with D k =3

16 CMSM achieves 100% throughput under non pre- emptive scheduling, if the traffic is constrained to less than cells for any input/output in B timeslots.  This introduces deterministic constraints on the arrival traffic  We are interested in the traditional stochastic traffic Previous Results Ref: [Weller, Hajek ], “Scheduling non-uniform traffic in a packet-switching system with small propagation delay,” IEEE/ACM Transactions on Networking 5(6): , 1997.

17 Arrival Traffic

18 Contents 1. Background & Motivation 2. Non Pre-emptive Scheduling 3. Achieving 100% throughput with CMSM Bernoulli i.i.d. uniform traffic Bernoulli i.i.d. non-uniform traffic 4. Achieving 100% throughput with MSM Bernoulli i.i.d. uniform traffic 5. Conclusion

19 CMSM with Uniform Traffic  Theorem 1: CMSM gives 100% throughput under non pre-emptive scheduling, if the input traffic is admissible and Bernoulli i.i.d. uniform  Informal Arguments:  Let T k be the time to schedule batch k  Then for batch k+1 we buffer new arrivals for time T k  We expect about  T k packets at every input/output  Hence, the maximum degree of batch k +1, i.e. D k+1   T k  Hence for a CMSM, T k+1 = D k+1   T k < T k  Hence T k is bounded in mean.

20  We are going to show that  Alternatively we will first show that  Observe that Formal Arguments Outline

21  We shall use the Chernoff bound to get  If we want to bound D k+1, we require that all the 2N vertices are bounded Formal Arguments … 1 Bounding the degree of a batch

22  Choose  > 0, such that.  Choose  such that  We get Formal Arguments … 2 Bounding the deviation of the service time of a batch

23  Hence Formal Arguments … 3 Bounding the service time of a batch

24  Choose  < (1-  ) /2,  This gives  Observe that  Q is now a function of T k only for a constant   We can make Q as close to 1, by choosing a large T k Formal Arguments …4 Tightening the bound

25  Hence, there is a constant T c such that  Formally, using a linear Lyapunov function V(T k ) = T k, we can say that T k (averaged over the batch index) is bounded in mean. Formal Arguments …5 Finishing Off..

26  In the paper we use a quadratic Lyapunov function V(T k ) = (T k ) 2, and show that T k 2 (averaged over the batch index) is bounded in mean.  There are a few technical steps after this to show that the queue size (averaged over time) is bounded in mean.  Then, it follows that CMSM gives 100% throughput for Bernoulli i.i.d. uniform traffic. Formal Arguments …6 Some Final Points..

27 Contents 1. Background & Motivation 2. Non Pre-emptive Scheduling 3. Achieving 100% throughput with CMSM Bernoulli i.i.d. uniform traffic Bernoulli i.i.d. non-uniform traffic 4. Achieving 100% throughput with MSM Bernoulli i.i.d. uniform traffic 5. Conclusion

28 CMSM with Non-Uniform Traffic  Theorem 2: CMSM achieves 100% throughput under non pre-emptive scheduling, if the input traffic is admissible and Bernoulli i.i.d.

29 Contents 1. Background & Motivation 2. Non Pre-emptive Scheduling 3. Achieving 100% throughput with CMSM Bernoulli i.i.d. uniform traffic Bernoulli i.i.d. non-uniform traffic 4. Achieving 100% throughput with MSM Bernoulli i.i.d. uniform traffic 5. Conclusion

30 Example of a Uniform Graph Batch Request Graph with D k =

31 MSM with Non-Uniform Traffic  Theorem 3: MSM achieves 100% throughput under non pre-emptive scheduling, if the input traffic is admissible and Bernoulli i.i.d. uniform

32 Contents 1. Background & Motivation 2. Non Pre-emptive Scheduling 3. Achieving 100% throughput with CMSM Bernoulli i.i.d. uniform traffic Bernoulli i.i.d. non-uniform traffic 4. Achieving 100% throughput with MSM Bernoulli i.i.d. uniform traffic 5. Conclusion

33 Conclusions  We have used the more traditional stochastic arrivals and shown using batch scheduling that  CMSM gives 100% throughput for Bernoulli i.i.d. traffic  MSM gives 100% throughput for Bernoulli i.i.d. uniform traffic  It would be nice to understand the stability of MSM with uniform load with continuous scheduling.