1. PCP: Efficient Endpoint Congestion Control NSDI, 2006 Thomas Anderson, Andrew Collins, Arvind Krishnamurthy and John Zahorjan University of Washington Presented by Aleksandar Kuzmanovic September 30, 2009

2. PCP -- Probe Control Protocol  Probe  Detect whether the network can currently support a test rate  End-to-end approach  Emulates network-based control  “Request and Set” Overview

3. Background TCP, Vegas, RAP, Fast TCP, S-TCP High Speed TCP DecBit, ECN, RED, AQM PCP ATM, XCP, WFQ, RCP End PointRouter Support Try and Backoff Request and Set

4. 1. Minimize transfer time 2. Negligible packet loss & low queue variability 3. Resources are fully allocated if there is sufficient demand 4. Fairness 5. Stable system even under high loads Design Goals

5. 1. Minimize transfer time Design Goals Common Case -- Most network paths are idle most of the time.  Most transfers are relatively short  Startup efficiency is particularly important.  TCP congestion control was designed at a time when links were thin and usually fully utilized  Efficiency loss of slow start is minimal

6. 2.Negligible packet loss & low queue variability Design Goals  Packet loss: Queue overflow  Can we prevent queues from overflow ?  Large queuing delays unnecessarily delay interactive response time and disrupt real-time traffic.  Can we eliminate queues that might build up at routers?

7. 1. Minimize transfer time 2. Negligible packet loss & low queue variability 3. Resources are fully allocated if there is sufficient demand 4. Fairness 5. Stable system even under high loads Design Goals Goals of PCP: Achieves rapid startup, small queues, and low loss rates, and that it does not compromise eventual efficiency, fairness and stability.

8.  Moderate sized flows on idle links  Interactive applications  Applications demanding minimally variable response times TCP managed networks perform poorly for these applications! Application Examples

9.  Test a target rate by sending a short probe.  Given a successful test, senders immediately increase their base rate by the target rate of the probe.  Two important techniques: Probe control: how to vary the test rates? Using history: achieves constant startup time Goal 1. Minimize transfer time  Direct Jump

10. Exponential increase and decrease  Start with a baseline rate: One maximum sized packet per round-trip.  Double the attempted rate increase after each successful probe.  Halve the attempted rate increase after each unsuccessful probe. Probe Control Time Rate Probe Channel Capacity Probe Direct Jump

11.  Send packet train spaced at an interval to achieve desired rate -- Currently, five packets whose size could be varied  Check for queuing delays based on reception times Probes

12. Using History Keep history information about the base rates previously used to each Internet address Set the initial probe rate based on previous base rate. Allows the end host to usually identify the optimal rate within two round trip times. Direct Jump

13. Goal 2. Negligible packet loss & low queue variability  Rate compensation Eliminate queues at routers:  Notice queue-buildups: Reduce the sending rate by a factor of (Δout – Δin ) /Δout  Detect persistent queueing: Reduce the sending rate by a factor of (max-delay – min-delay) / max-delay

14.  Transmit the baseline packets in a paced manner (equally spaced) at the base rate.  Monitor the gap between baseline PCP packets Δin -- gap used by the sender Δout -- gap observed at the receiver  Monitor the one-way delays of baseline PCP packets max-delay -- maximum one-way delay (maxdelay) observed in the previous round trip time min-delay -- minimum observed one-way delay (will time out) Baseline Packets

15.  Send packet train spaced at an interval to achieve desired rate -- Currently, five packets whose size could be varied  Check for queuing delays based on reception times Probes

16. Both are paced packets. Probes: short, high-rate bursts (sent at a test rate) Baseline packets: regular data traffic (sent at the base rate) Impact of a Probe is independent of its test rate. Easy to test aggressively without fear of disrupting existing connections. Comparison of Baseline Packets & Probes Time Rate Probe Channel Capacity Probe

17.  User-level implementation: Response time improves by a factor of 2 over TCP Better performance for long transfers as well  Simulation Smaller response times, smaller queue sizes PCP is compatible and benefits from fair queuing in the network Evaluation

18. Conclusion Emulating ‘request and set’ approach from the endpoints is possible The key idea is to send short probes: - impact of a probe is independent of its test rate - easy to test aggressively Achieves negligible packet loss & low queue variability Superior start-up behavior Effective over FQ routers