Mohammad Alizadeh Adel Javanmard and Balaji Prabhakar Stanford University Analysis of DCTCP:Analysis of DCTCP: Stability, Convergence, and FairnessStability,

Slides:



Advertisements
Similar presentations
Balaji Prabhakar Active queue management and bandwidth partitioning algorithms Balaji Prabhakar Departments of EE and CS Stanford University
Advertisements

Mohammad Alizadeh, Albert Greenberg, David A. Maltz, Jitendra Padhye Parveen Patel, Balaji Prabhakar, Sudipta Sengupta, Murari Sridharan Microsoft Research.
FAST TCP Anwis Das Ajay Gulati Slides adapted from : IETF presentation slides Link:
Deconstructing Datacenter Packet Transport Mohammad Alizadeh, Shuang Yang, Sachin Katti, Nick McKeown, Balaji Prabhakar, Scott Shenker Stanford University.
Mohammad Alizadeh, Albert Greenberg, David A. Maltz, Jitendra Padhye Parveen Patel, Balaji Prabhakar, Sudipta Sengupta, Murari Sridharan Modified by Feng.
Lecture 18: Congestion Control in Data Center Networks 1.
Mohammad Alizadeh, Albert Greenberg, David A. Maltz, Jitendra Padhye Parveen Patel, Balaji Prabhakar, Sudipta Sengupta, Murari Sridharan Presented by Shaddi.
Fixing TCP in Datacenters Costin Raiciu Advanced Topics in Distributed Systems 2011.
PFabric: Minimal Near-Optimal Datacenter Transport Mohammad Alizadeh Shuang Yang, Milad Sharif, Sachin Katti, Nick McKeown, Balaji Prabhakar, Scott Shenker.
Congestion Control: TCP & DC-TCP Swarun Kumar With Slides From: Prof. Katabi, Alizadeh et al.
Microscopic Behavior of Internet Control Xiaoliang (David) Wei NetLab, CS&EE California Institute of Technology.
Balajee Vamanan et al. Deadline-Aware Datacenter TCP (D 2 TCP) Balajee Vamanan, Jahangir Hasan, and T. N. Vijaykumar.
5/17/20151 Adaptive RED: An Algorithm for Increasing the Robustness of RED’s Active Queue Management or How I learned to stop worrying and love RED Presented.
Advanced Computer Networking Congestion Control for High Bandwidth-Delay Product Environments (XCP Algorithm) 1.
Congestion Control An Overview -Jyothi Guntaka. Congestion  What is congestion ?  The aggregate demand for network resources exceeds the available capacity.
XCP: Congestion Control for High Bandwidth-Delay Product Network Dina Katabi, Mark Handley and Charlie Rohrs Presented by Ao-Jan Su.
TCP Stability and Resource Allocation: Part II. Issues with TCP Round-trip bias Instability under large bandwidth-delay product Transient performance.
On Modeling Feedback Congestion Control Mechanism of TCP using Fluid Flow Approximation and Queuing Theory  Hisamatu Hiroyuki Department of Infomatics.
One More Bit Is Enough Yong Xia, RPI Lakshminarayanan Subramanian, UCB Ion Stoica, UCB Shivkumar Kalyanaraman, RPI SIGCOMM’05, August 22-26, 2005, Philadelphia,
Bertha & M Sadeeq.  Easy to manage the problems  Scalability  Real time and real environment  Free data collection  Cost efficient  DCTCP only covers.
Congestion control in data centers
AQM for Congestion Control1 A Study of Active Queue Management for Congestion Control Victor Firoiu Marty Borden.
Defense: Christopher Francis, Rumou duan Data Center TCP (DCTCP) 1.
SEDCL: Stanford Experimental Data Center Laboratory.
1 Random Early Detection Gateways for Congestion Avoidance Sally Floyd and Van Jacobson, IEEE Transactions on Networking, Vol.1, No. 4, (Aug 1993), pp
Modeling TCP in Small-Buffer Networks
Fluid-based Analysis of a Network of AQM Routers Supporting TCP Flows with an Application to RED Vishal Misra Wei-Bo Gong Don Towsley University of Massachusetts,
1 Emulating AQM from End Hosts Presenters: Syed Zaidi Ivor Rodrigues.
Computer Networking Lecture 17 – Queue Management As usual: Thanks to Srini Seshan and Dave Anderson.
Congestion Control for High Bandwidth-delay Product Networks Dina Katabi, Mark Handley, Charlie Rohrs.
CS :: Fall 2003 TCP Friendly Streaming Ketan Mayer-Patel.
Congestion Control for High Bandwidth-Delay Product Environments Dina Katabi Mark Handley Charlie Rohrs.
1 A State Feedback Control Approach to Stabilizing Queues for ECN- Enabled TCP Connections Yuan Gao and Jennifer Hou IEEE INFOCOM 2003, San Francisco,
Mohammad Alizadeh, Abdul Kabbani, Tom Edsall,
Mohammad Alizadeh Stanford University Joint with: Abdul Kabbani, Tom Edsall, Balaji Prabhakar, Amin Vahdat, Masato Yasuda HULL: High bandwidth, Ultra Low-Latency.
Balaji Prabhakar Mohammad Alizadeh, Abdul Kabbani, and Berk Atikoglu Stanford University Stability Analysis of QCN:Stability Analysis of QCN: The Averaging.
Mohammad Alizadeh, Albert Greenberg, David A. Maltz, Jitendra Padhye Parveen Patel, Balaji Prabhakar, Sudipta Sengupta, Murari Sridharan by liyong Data.
TCP & Data Center Networking
Curbing Delays in Datacenters: Need Time to Save Time? Mohammad Alizadeh Sachin Katti, Balaji Prabhakar Insieme Networks Stanford University 1.
Congestion models for bursty TCP traffic Damon Wischik + Mark Handley University College London DARPA grant W911NF
TFRC: TCP Friendly Rate Control using TCP Equation Based Congestion Model CS 218 W 2003 Oct 29, 2003.
Understanding the Performance of TCP Pacing Amit Aggarwal, Stefan Savage, Thomas Anderson Department of Computer Science and Engineering University of.
B 李奕德.  Abstract  Intro  ECN in DCTCP  TDCTCP  Performance evaluation  conclusion.
Advance Computer Networking L-6 TCP & Routers Acknowledgments: Lecture slides are from the graduate level Computer Networks course thought by Srinivasan.
Fluid-based Analysis of a Network of AQM Routers Supporting TCP Flows with an Application to RED Vishal Misra Wei-Bo Gong Don Towsley University of Massachusetts,
ACN: RED paper1 Random Early Detection Gateways for Congestion Avoidance Sally Floyd and Van Jacobson, IEEE Transactions on Networking, Vol.1, No. 4, (Aug.
Packet Transport Mechanisms for Data Center Networks Mohammad Alizadeh NetSeminar (April 12, 2012) Mohammad Alizadeh NetSeminar (April 12, 2012) Stanford.
MaxNet NetLab Presentation Hailey Lam Outline MaxNet as an alternative to TCP Linux implementation of MaxNet Demonstration of fairness, quick.
27th, Nov 2001 GLOBECOM /16 Analysis of Dynamic Behaviors of Many TCP Connections Sharing Tail-Drop / RED Routers Go Hasegawa Osaka University, Japan.
Congestion control for Multipath TCP (MPTCP) Damon Wischik Costin Raiciu Adam Greenhalgh Mark Handley THE ROYAL SOCIETY.
DCTCP & CoDel the Best is the Friend of the Good Bob Briscoe, BT Murari Sridharan, Microsoft IETF-84 TSVAREA Jul 2012 Le mieux est l'ennemi du bien Voltaire.
1 IEEE Meeting July 19, 2006 Raj Jain Modeling of BCN V2.0 Jinjing Jiang and Raj Jain Washington University in Saint Louis Saint Louis, MO
Acknowledgments S. Athuraliya, D. Lapsley, V. Li, Q. Yin (UMelb) S. Adlakha (UCLA), J. Doyle (Caltech), K. Kim (SNU/Caltech), F. Paganini (UCLA), J. Wang.
1 Analysis of a window-based flow control mechanism based on TCP Vegas in heterogeneous network environment Hiroyuki Ohsaki Cybermedia Center, Osaka University,
HP Labs 1 IEEE Infocom 2003 End-to-End Congestion Control for InfiniBand Jose Renato Santos, Yoshio Turner, John Janakiraman HP Labs.
1 Sheer volume and dynamic nature of video stresses network resources PIE: A lightweight latency control to address the buffer problem issue Rong Pan,
6.888: Lecture 3 Data Center Congestion Control Mohammad Alizadeh Spring
Mohammad Alizadeh, Albert Greenberg, David A. Maltz, Jitendra Padhye Parveen Patel, Balaji Prabhakar, Sudipta Sengupta, Murari Sridharan Microsoft Research.
Data Center TCP (DCTCP)
Data Center TCP (DCTCP)
CS 268: Lecture 6 Scott Shenker and Ion Stoica
Router-Assisted Congestion Control
Packet Transport Mechanisms for Data Center Networks
Columbia University in the city of New York
Microsoft Research Stanford University
Data Center TCP (DCTCP)
Lecture 16, Computer Networks (198:552)
Lecture 17, Computer Networks (198:552)
Understanding Congestion Control Mohammad Alizadeh Fall 2018
Adaptive RED: An Algorithm for Increasing the Robustness of RED’s Active Queue Management or How I learned to stop worrying and love RED Presented by:
Presentation transcript:

Mohammad Alizadeh Adel Javanmard and Balaji Prabhakar Stanford University Analysis of DCTCP:Analysis of DCTCP: Stability, Convergence, and FairnessStability, Convergence, and Fairness

Data Center Packet TransportData Center Packet Transport Transport inside the DC – TCP rules (99.9% of traffic in some DCs) But, TCP: – Needs large buffers for high throughput – Induces large queuing delays – Does not handle bursty traffic well (Incast) DCTCP was proposed to address these shortcomings (SIGCOMM’10). 2

TCP Buffer RequirementTCP Buffer Requirement Bandwidth-delay product rule of thumb: – A single flow needs C×RTT buffers for 100% Throughput. B Buffer Size B = C×RTT B B < C×RTT Buffer Size Throughput loss! B Buffer Size B > C×RTT More latency! To lower the buffering requirements, we must reduce sending rate variations. 3

DCTCP: Main IdeasDCTCP: Main Ideas 1.React in proportion to the extent of congestion. Reduce window size based on fraction of marked packets. 2.Mark based on instantaneous queue length. Fast feedback to better deal with bursts. Simplifies hardware. ECN MarksTCPDCTCP Cut window by 50%Cut window by 40% Cut window by 50%Cut window by 5% 4

DCTCP: AlgorithmDCTCP: Algorithm Switch side: – Mark packets when Queue Length > K. Sender side: – Maintain running average of fraction of packets marked (α).  Adaptive window decreases: – Note: decrease factor between 1 and 2. B K Mark Don’t Mark 5

DCTCP vs TCPDCTCP vs TCP Setup: Win 7, Broadcom 1Gbps Switch Scenario: 2 long-lived flows, K = 30KB (Kbytes) 6

Analysis of DCTCPAnalysis of DCTCP

Steady State AnalysisSteady State Analysis What is the effect of the various network and algorithm parameters on system throughput and latency? – Network: Capacity, Round-trip Time, Number of flows – Algorithm: Marking threshold (K), Averaging parameter (g) The standard approach is to study control loop behavior via fluid models. – Kelly et al., Low et al., Misra et al., Srikant et al, … 8

DCTCP Fluid ModelDCTCP Fluid Model 9 × N/RTT(t) W(t) p(t) Delay p(t – R * ) C + − 1 0 K q(t) Switch LPF AIMD α(t) Source

Fluid Model vs ns2 simulationsFluid Model vs ns2 simulations Parameters: N = {2, 10, 100}, C = 10Gbps, d = 100μs, K = 65 pkts, g = 1/16. N = 2N = 10N =

We make the following change of variables: The normalized system: The normalized system depends on only two parameters: Normalization of Fluid ModelNormalization of Fluid Model 11

Equilibrium Characterization Case 1: Equilibrium Characterization Case 1: Very large N: system (globally) converges to a unique fixed point: Example: 12

Very large N: system (globally) converges to a unique fixed point: 12 Example: Equilibrium Characterization Case 1: Equilibrium Characterization Case 1:

System has a periodic limit cycle solution. Example: 13 Equilibrium Characterization Case 2: Equilibrium Characterization Case 2:

System has a periodic limit cycle solution. Example: 13 Equilibrium Characterization Case 2: Equilibrium Characterization Case 2:

Stability of Limit CyclesStability of Limit Cycles Let X * = set of points on the limit cycle. A limit cycle is locally asymptotically stable if δ > 0 exists s.t.: 14

Poincaré MapPoincaré Map 15 x1x1 x2x2 x 2 = P(x 1 ) Stability of Poincaré Map ↔ Stability of limit cycle x * α = P(x * α )

Stability CriterionStability Criterion Theorem: The limit cycle of the DCTCP system: is locally asymptotically stable if and only if ρ(Z 1 Z 2 ) < 1. -J F is the Jacobian matrix with respect to x. -T = (1 + h α )+(1 + h β ) is the period of the limit cycle. Proof: Show that P(x * α + δ) = x * α + Z 1 Z 2 δ + O(|δ| 2 ). 16 We have numerically checked this condition for:

Parameter GuidelinesParameter Guidelines How big does the marking threshold K need to be to avoid queue underflow? B K 17

Throughput-Latency TradeoffThroughput-Latency Tradeoff Throughput > 94% as K  0 18 Parameters: C = 10Gbps, d = 480μs, g = For TCP: Throughput → 75% For TCP: Throughput → 75%

Convergence AnalysisConvergence Analysis How long does it take for DCTCP sources to converge to their “fair share” rate (C/N)? – DCTCP is slower to converge than TCP since it cuts its window by smaller factors. The fluid model is not suitable for transient analyses. We use a hybrid (continuous- and discrete-time) model. – The model is inspired by the AIMD models of Baccelli et al. and Shorten et al. 19

The Hybrid ModelThe Hybrid Model 20 Time Window Sizes Time p(t) (Marking Prob.) 1 RTT

Rate of Convergence (Theorem)Rate of Convergence (Theorem) Assume N DCTCP flows with arbitrary W i (0) and α i (0), evolving according to the Hybrid Model, with: Define function, and let 0 < α * ≤ 1 be the unique positive solution to Then: Also: where: 21

Consequences DCTCP converges at most 40% slower than TCP: The parameter g should not be too small: 22

(g = 0.07) (g = 0.025) (g = 0.005) Convergence: ns2 SimulationsConvergence: ns2 Simulations 23

Conclusion Our analysis shows DCTCP: – requires 17% of C×RTT for full throughput – achieves 94% throughput as K → 0. – converges at most 1.4 times slower than TCP. We provide guidelines for setting the DCTCP parameters. The analysis suggests a simple modification that improves the RTT-fairness of DCTCP. – Achieves linear-RTT fairness (Thrput RTT -1 ), like TCP-RED 24

25