1 / 21

FAST TCP: From Theory to Experiments

FAST TCP: From Theory to Experiments. C. Jin, D. Wei, S. H. Low, G. Buhrmaster, J. Bunn, D. H. Choe, R. L. A. Cottrell, J. C. Doyle, W. Feng, O. Martin, H. Newman, F. Paganini, S. Ravot, S. Singh. Motivation. High Energy and Nuclear Physics (HENP)

masato
Télécharger la présentation

FAST TCP: From Theory to Experiments

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. FAST TCP: From Theory to Experiments C. Jin, D. Wei, S. H. Low, G. Buhrmaster, J. Bunn, D. H. Choe, R. L. A. Cottrell, J. C. Doyle, W. Feng, O. Martin, H. Newman, F. Paganini, S. Ravot, S. Singh Anshul Kantawala

  2. Motivation • High Energy and Nuclear Physics (HENP) • 2,000 physicists, 150 universities, 30 countries • Require ability to analyze and share many terabyte-scale data collections Key Challenge Current congestion control algorithm of TCP does not scale to this regime Anshul Kantawala

  3. Background Theory • Congestion control consists of: • Source algorithm (TCP) that adapts sending rate (window) based on congestion feedback • Link algorithm (at router) that updates and feeds back a measure of congestion • Typically link algorithm is implicit and measure of congestion is loss probability or queueing delay Anshul Kantawala

  4. Background Theory (cont.) • Preliminary theory studies equilibrium and stability properties of the source-link algorithm pair • Source-link algorithm is TCP/AQM (active queue management) Anshul Kantawala

  5. Equilibrium • Interpret TCP/AQM as a distributed algorithm to solve a global optimization problem • Can be broken into two sub-problems • Source (maximize sum of all data rates) • Link (minimize congestion) Anshul Kantawala

  6. Equilibrium (cont.) • Source • Each source has a utility function, as a function of its data rate • Optimization problem is maximizing sum of all utility functions over their rates • Challenge is solving for optimal source rates in a distributed manner using only local information Anshul Kantawala

  7. Equilibrium (cont.) • Exploit Duality Theory • Associated with primal (source) utility maximization is dual (link congestion) minimization problem • Solving the dual problem is equivalent to solving the primal problem • Class of optimization algorithms that iteratively solve for both at once Anshul Kantawala

  8. Stability • Want to ensure equilibrium points are stable • When the network is perturbed out of equilibrium by random fluctuations, should drift back to new equilibrium point • Current TCP algorithms can become unstable as delay increases or network capacity increases Anshul Kantawala

  9. ns-2 simulation Lack of Scalability of TCP • As capacity increases, link utilization of TCP/RED steadily drops • Main factor may be synchronization of TCP flows capacity = 155Mbps, 622Mbps, 2.5Gbps, 5Gbps, 10Gbps; 100 ms round trip latency; 100 flows Anshul Kantawala

  10. FAST TCP • Implementation issues • Use both queueing delay and packet loss as congestion signals • Effectively deal with massive losses • Use pacing to reduce burstiness and massive losses • Converge rapidly to neighbourhood of equilibrium value after packet losses Anshul Kantawala

  11. FAST TCP - Implementation • Congestion window update (every RTT) Wnew = 1/2( [Wold * baseRTT/avgRTT] +  + Wcurrent) Anshul Kantawala

  12. Calculating parameters • RTT and Wold • Timestamp every packet stored in retransmit queue and store Wcurrent • On receipt of ACK, set RTT sample to difference in times. Also, set Wold to the stored window size of the ACK’ed packet • baseRTT is set to minimum observed RTT sample Anshul Kantawala

  13. Calculating parameters (cont.) • avgRTT avgRTTnew = (1-weight) * avgRTTold + weight * RTT where weight = min(3/cwnd, 1/8) • With a large window, successive RTT samples capture congestion information at a time scale much smaller than RTT Anshul Kantawala

  14. Calculating parameters (cont.) •  • Specifies total number of packets a single FAST connection tries to maintain in queues along its path • Can be used to tune the aggressiveness of the window update function Anshul Kantawala

  15. Experiments - Infrastructure Anshul Kantawala

  16. Infrastructure - Details 7, 9, 10 FAST flows 3,948 km 1, 2 Linux/FAST flows 10,037 km Anshul Kantawala

  17. Throughput and Utilization • Statistics in parentheses are for current Linux TCP implementation • “bmps” = product of throughput and distance (bits-per-meter-per-second) Anshul Kantawala

  18. 92% 2G 48% Average utilization 95% 1G 27% 16% 19% Linux TCP Linux TCP FAST Linux TCP Linux TCP FAST Average Utilization Traces Anshul Kantawala

  19. 88% 90% 90% Average utilization 92% 95% 1 flow 2 flows 7 flows 9 flows 10 flows Traces (cont.) Anshul Kantawala

  20. Fairness • 1 flow from CERN to Sunnyvale and 2 flows from CERN to Chicago Anshul Kantawala

  21. Fairness (cont.) • Average throughputs • Single FAST flow achieved 760 Mbps, 2 Linux flows achieved 648 Mbps and 446 Mbps respectively • Single Linux flow achieved 419 Mbps, 2 FAST flows achieved 841 Mbps and 732 Mbps respectively Anshul Kantawala

More Related