115-744: Computer NetworkingL-16 Multicast Reliability and Congestion ControlL -16; 11-5-02© Srinivasan Seshan, 2002 2Multicast Issues• Reliable multicast• Multicast congestion control• Assigned reading• [F+97] A Reliable Multicast Framework for Light-Weight Sessions and Application Level FramingL -16; 11-5-02© Srinivasan Seshan, 2002 3Overview• Scalable Reliable Multicast• Congestion ControlL -16; 11-5-02© Srinivasan Seshan, 2002 4SRM• Originally designed for wb• Receiver-reliable• NACK-based• Every member may multicast NACK or retransmission2L -16; 11-5-02© Srinivasan Seshan, 2002 5R1SRM Request SuppressionSR3R221R1SR3R2Packet 1 is lost; R1 requests resend to Source and ReceiversPacket 1 is resent; R2 and R3 no longer have to request a resend1XXDelay varies by distanceXResend requestResent packetL -16; 11-5-02© Srinivasan Seshan, 2002 6Deterministic Suppressiondddd3dTimedatanack repaird4dd2d3d= Sender= Repairer= RequestorDelay = C1×dS,R L -16; 11-5-02© Srinivasan Seshan, 2002 7SRM Star TopologySR221R3Packet 1 is lost; All Receivers request resendsPacket 1 is resent to all ReceiversXR4Delay is same lengthSR21R3 R4Resend requestResent packetL -16; 11-5-02© Srinivasan Seshan, 2002 8SRM: Stochastic SuppressiondataddddTimeNACKrepair2dsession msg0123Delay = U[0,D2] ×dS,R= Sender= Repairer= Requestor3L -16; 11-5-02© Srinivasan Seshan, 2002 9SRM (Summary)• NACK/Retransmission suppression• Delay before sending• Delay based on RTT estimation• Deterministic + Stochastic components• Periodic session messages• Full reliability• Estimation of distance matrix among membersL -16; 11-5-02© Srinivasan Seshan, 2002 10What’s Missing?• Losses at link (A,C) causes retransmission to the whole group• Only retransmit to those members who lost the packet• [Only request from the nearest responder]SenderReceiverA BE FSC D0.990 00 0 0L -16; 11-5-02© Srinivasan Seshan, 2002 11Local Recovery• Different techniques in various systems• Application-level hierarchy• Fixed v.s. dynamic• TTL scoped multicast• Router supportedL -16; 11-5-02© Srinivasan Seshan, 2002 12Overview• Scalable Reliable Multicast• Congestion Control4L -16; 11-5-02© Srinivasan Seshan, 2002 13Multicast Congestion Control• What if receivers have very different bandwidths?• Send at max?• Send at min?• Send at avg?RRRS???Mb/s100Mb/s100Mb/s1Mb/s1Mb/s56Kb/sRL -16; 11-5-02© Srinivasan Seshan, 2002 14Video Adaptation: RLM• Receiver-driven Layered Multicast• Layered video encoding• Each layer uses its own mcast group• On spare capacity, receivers add a layer• On congestion, receivers drop a layer• Join experiments used for shared learningL -16; 11-5-02© Srinivasan Seshan, 2002 15Layered Media StreamsSRR1R2R3R10Mbps10Mbps512Kbps128Kbps10MbpsR3 joins layer 1, fails at layer 2R2 join layer 1,join layer 2 fails at layer 3R1 joins layer 1,joins layer 2 joins layer 3L -16; 11-5-02© Srinivasan Seshan, 2002 16Drop Policies for Layered Multicast• Priority• Packets for low bandwidth layers are kept, drop queued packets for higher layers• Requires router support • Uniform (e.g., drop tail, RED)• Packets arriving at congested router are dropped regardless of their layer• Which is better?• Intuition vs. reality!5L -16; 11-5-02© Srinivasan Seshan, 2002 17RLM IntuitionUniform vs. Priority Dropping010203040506070Offered loadPerformanceUniformPriorityL -16; 11-5-02© Srinivasan Seshan, 2002 18RLM Intuition• Uniform• Better incentives to well-behaved users• If oversend, performance rapidly degrades• Clearer congestion signal• Allows shared learning• Priority• Can waste upstream resources• Hard to deploy• RLM approaches optimal operating point• Uniform is already deployed• Assume no special router supportL -16; 11-5-02© Srinivasan Seshan, 2002 19RLM Join Experiment• Receivers periodically try subscribing to higher layer• If enough capacity, no congestion, no drops à Keep layer (& try next layer)• If not enough capacity, congestion, drops à Drop layer (& increase time to next retry)• What about impact on other receivers?L -16; 11-5-02© Srinivasan Seshan, 2002 20Join Experiments1234TimeLayer6L -16; 11-5-02© Srinivasan Seshan, 2002 21RLM Scalability?• What happens with more receivers?• Increased frequency of experiments?• More likely to conflict (false signals)• Network spends more time congested• Reduce # of experiments per host?• Takes longer to converge• Receivers coordinate to improve behaviorL -16; 11-5-02© Srinivasan Seshan, 2002 22Next Lecture: Security• Denial of service• IPSec• Firewalls• Assigned reading• [SWKA00] Practical Network Support for IP Traceback• [B89] Security Problems in the TCP/IP Protocol
View Full Document
Unlocking...