Fall 2006CS 5611 Congestion Control Outline Queuing Discipline Reacting to Congestion Avoiding...
-
date post
19-Dec-2015 -
Category
Documents
-
view
218 -
download
3
Transcript of Fall 2006CS 5611 Congestion Control Outline Queuing Discipline Reacting to Congestion Avoiding...
Fall 2006 CS 561 1
Congestion Control
OutlineQueuing Discipline
Reacting to Congestion
Avoiding Congestion
Quality of Service
Fall 2006 CS 561 2
Issues• Two sides of the same coin
– pre-allocate resources so at to avoid congestion– control congestion if (and when) is occurs
• Two points of implementation– hosts at the edges of the network (transport protocol)– routers inside the network (queuing discipline)
• Underlying service model– best-effort– multiple qualities of service (QoS)
Destination1.5-Mbps T1 link
Router
Source2
Source1
100-Mbps FDDI
10-Mbps Ethernet
Fall 2006 CS 561 3
Framework• Connectionless flows
– sequence of packets sent between source/destination pair– maintain soft state at the routers
• Taxonomy– router-centric versus host-centric– reservation-based versus feedback-based– window-based versus rate-based
Router
Source2
Source1
Source3
Router
Router
Destination2
Destination1
Fall 2006 CS 561 4
Evaluation
• Fairness• Power (ratio of throughput to delay)
Optimalload Load
Th
rou
ghp
ut/d
elay
Fall 2006 CS 561 5
Queuing Discipline• First-In-First-Out (FIFO)
– does not discriminate between traffic sources– drop policy (tail-drop, random early drop)
• Fair Queuing (FQ)– explicitly segregates traffic based on flows– ensures no flow captures more than its share of capacity– variation: weighted fair queuing (WFQ)
• Problem?Flow 1
Flow 2
Flow 3
Flow 4
Round-robinservice
Fall 2006 CS 561 6
FQ Algorithm
• Suppose clock ticks each time a bit is transmitted• Let Pi denote the length of packet i• Let Si denote the time when start to transmit packet i• Let Fi denote the time when finish transmitting packet i• Fi = Si + Pi
• When does router start transmitting packet i?– if before router finished packet i - 1 from this flow, then
immediately after last bit of i - 1 (Fi-1)– if no current packets for this flow, then start transmitting
when arrives (call this Ai)• Thus: Fi = MAX (Fi - 1, Ai) + Pi
Fall 2006 CS 561 7
FQ Algorithm (cont)
• For multiple flows– calculate Fi for each packet that arrives on each flow– treat all Fi’s as timestamps– next packet to transmit is one with lowest timestamp
• Not perfect: can’t preempt current packet• Example
Flow 1 Flow 2
(a) (b)
Output Output
F = 8 F = 10F = 5
F = 10
F = 2
Flow 1(arriving)
Flow 2(transmitting)
Fall 2006 CS 561 8
TCP Congestion Control
• Idea– assumes best-effort network (FIFO or FQ routers) each
source determines network capacity for itself
– uses implicit feedback
– ACKs pace transmission (self-clocking)
• Challenge– determining the available capacity in the first place
– adjusting to changes in the available capacity
Fall 2006 CS 561 9
Additive Increase/Multiplicative Decrease
• Objective: adjust to changes in the available capacity• New state variable per connection: CongestionWindow
– limits how much data source has in transit
MaxWin = MIN(CongestionWindow, AdvertisedWindow)
EffWin = MaxWin - (LastByteSent - LastByteAcked)
• Idea:– increase CongestionWindow when congestion goes down– decrease CongestionWindow when congestion goes up
Fall 2006 CS 561 10
AIMD (cont)
• Question: how does the source determine whether or not the network is congested?
• Answer: a timeout occurs– timeout signals that a packet was lost– packets are seldom lost due to transmission error– lost packet implies congestion
Fall 2006 CS 561 11
AIMD (cont)
• In practice: increment a little for each ACKIncrement = (MSS * MSS)/CongestionWindow
CongestionWindow += Increment
Source Destination
…
• Algorithm– increment CongestionWindow by
one packet per RTT (linear increase)
– divide CongestionWindow by two whenever a timeout occurs (multiplicative decrease)
Fall 2006 CS 561 12
AIMD (cont)
• Trace: sawtooth behavior
60
20
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
KB
Time (seconds)
70
304050
10
10.0
Fall 2006 CS 561 13
Slow Start
• Objective: determine the available capacity in the first place
• Idea:– begin with CongestionWindow = 1
packet– double CongestionWindow each RTT
(increment by 1 packet for each ACK)
Source Destination
…
Fall 2006 CS 561 14
Slow Start (cont)• Exponential growth, but slower than all at once• Used…
– when first starting connection– when connection goes dead waiting for timeout
• Trace
• Problem: lose up to half a CongestionWindow’s worth of data
60
20
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
KB
70
304050
10
Fall 2006 CS 561 15
Fast Retransmit and Fast Recovery
• Problem: coarse-grain TCP timeouts lead to idle periods
• Fast retransmit: use duplicate ACKs to trigger retransmission
Packet 1
Packet 2
Packet 3
Packet 4
Packet 5
Packet 6
Retransmitpacket 3
ACK 1
ACK 2
ACK 2
ACK 2
ACK 6
ACK 2
Sender Receiver
Fall 2006 CS 561 16
Results
• Fast recovery– skip the slow start phase– go directly to half the last successful CongestionWindow (ssthresh)
60
20
1.0 2.0 3.0 4.0 5.0 6.0 7.0
KB
70
304050
10
Fall 2006 CS 561 17
Congestion Avoidance• TCP’s strategy
– control congestion once it happens
– repeatedly increase load in an effort to find the point at which congestion occurs, and then back off
• Alternative strategy– predict when congestion is about to happen
– reduce rate before packets start being discarded
– call this congestion avoidance, instead of congestion control
• Two possibilities – router-centric: DECbit and RED Gateways
– host-centric: TCP Vegas
Fall 2006 CS 561 18
DECbit• Add binary congestion bit to each packet header• Router
– monitors average queue length over last busy+idle cycle
– set congestion bit if average queue length > 1– attempts to balance throughout against delay
Queue length
Currenttime
TimeCurrent
cyclePrevious
cycleAveraginginterval
Fall 2006 CS 561 19
DECbit (cont)
• Destination echoes bit back to source• Source records how many packets resulted in set bit• If less than 50% of last window’s worth had bit set
– increase CongestionWindow by 1 packet
• If 50% or more of last window’s worth had bit set – decrease CongestionWindow by 0.875 times
Fall 2006 CS 561 20
Random Early Detection (RED)
• Notification is implicit – just drop the packet (TCP will timeout)– could make explicit by marking the packet
• Early random drop– rather than wait for queue to become full, drop each
arriving packet with some drop probability whenever the queue length exceeds some drop level
Fall 2006 CS 561 21
RED Details• Compute average queue length
AvgLen = (1 - Weight) * AvgLen + Weight * SampleLen
0 < Weight < 1 (usually 0.002)SampleLen is queue length each time a packet arrives
MaxThreshold MinThreshold
AvgLen
Fall 2006 CS 561 22
RED Details (cont)
• Two queue length thresholds
if AvgLen <= MinThreshold then
enqueue the packet
if MinThreshold < AvgLen < MaxThreshold then
calculate probability P
drop arriving packet with probability P
if MaxThreshold <= AvgLen then
drop arriving packet
Fall 2006 CS 561 23
RED Details (cont)• Computing probability P
TempP = MaxP * (AvgLen - MinThreshold)/ (MaxThreshold - MinThreshold)
P = TempP/(1 - count * TempP)
• Drop Probability CurveP(drop)
1.0
MaxP
MinThresh MaxThresh
AvgLen
Fall 2006 CS 561 24
Tuning RED• Probability of dropping a particular flow’s packet(s) is roughly
proportional to the share of the bandwidth that flow is currently getting• MaxP is typically set to 0.02, meaning that when the average queue size
is halfway between the two thresholds, the gateway drops roughly one out of 50 packets.
• If traffic id bursty, then MinThreshold should be sufficiently large to allow link utilization to be maintained at an acceptably high level
• Difference between two thresholds should be larger than the typical increase in the calculated average queue length in one RTT; setting MaxThreshold to twice MinThreshold is reasonable for traffic on today’s Internet
• Penalty Box for Offenders
Fall 2006 CS 561 25
TCP Vegas• Idea: source watches for some sign that router’s queue is
building up and congestion will happen too; e.g.,– RTT grows
– sending rate flattens 60
20
0.5 1.0 1.5 4.0 4.5 6.5 8.0
KB
Time (seconds)
Time (seconds)
70
304050
10
2.0 2.5 3.0 3.5 5.0 5.5 6.0 7.0 7.5 8.5
900
300100
0.5 1.0 1.5 4.0 4.5 6.5 8.0
Sen
din
g K
Bp
s 1100
500700
2.0 2.5 3.0 3.5 5.0 5.5 6.0 7.0 7.5 8.5
Time (seconds)0.5 1.0 1.5 4.0 4.5 6.5 8.0Q
ueu
e s
ize
in r
oute
r
5
10
2.0 2.5 3.0 3.5 5.0 5.5 6.0 7.0 7.5 8.5
Fall 2006 CS 561 26
Algorithm • Let BaseRTT be the minimum of all measured RTTs (commonly the
RTT of the first packet)• If not overflowing the connection, then
ExpectRate = CongestionWindow/BaseRTT• Source calculates sending rate (ActualRate) once per RTT• Source compares ActualRate with ExpectRate
Diff = ExpectedRate - ActualRateif Diff <
increase CongestionWindow linearlyelse if Diff >
decrease CongestionWindow linearlyelse
leave CongestionWindow unchanged
Fall 2006 CS 561 27
Algorithm (cont)
• Parameters = 1 packet = 3 packets
• Even faster retransmit– keep fine-grained timestamps for each packet – check for timeout on first duplicate ACK
70605040302010
KB
Time (seconds)
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
CA
M K
Bps
2402001601208040
Time (seconds)
Fall 2006 CS 561 28
Realtime Applications• Require “deliver on time” assurances
– must come from inside the network
• Example application (audio)– sample voice once every 125us– each sample has a playback time– packets experience variable delay in network– add constant factor to playback time: playback point
Microphone
Speaker
Sampler,A D
converter
Buffer,D A
Fall 2006 CS 561 29
Playback BufferS
eque
nce
num
ber
Packetgeneration
Networkdelay
Buffer
Playback
Time
Packetarrival
Fall 2006 CS 561 30
Example Distribution of Delays
1
2
3
Pa
cke
ts (
%)
90% 97% 98% 99%
150 20010050
Delay (milliseconds)
Fall 2006 CS 561 31
Integrated Services
• Service Classes– guaranteed
– controlled-load
• Mechanisms– signalling protocol
– admission control
– policing
– packet scheduling
Fall 2006 CS 561 32
Flowspec
• Rspec: describes service requested from network– controlled-load: none– guaranteed: delay target
• Tspec: describes flow’s traffic characteristics– average bandwidth + burstiness: token bucket filter– token rate r– bucket depth B– must have a token to send a byte– must have n tokens to send n bytes– start with no tokens– accumulate tokens at rate of r per second– can accumulate no more than B tokens
Fall 2006 CS 561 33
Differentiated Services• Problem with IntServ: scalability• Idea: segregate packets into a small number of classes
– e.g., premium vs best-effort
• Packets marked according to class at edge of network• Core routers implement some per-hop-behavior (PHB)• Example: Expedited Forwarding (EF)
– rate-limit EF packets at the edges– PHB implemented with class-based priority queues or WFQ
Fall 2006 CS 561 34
DiffServ (cont)
• Assured Forwarding (AF)– customers sign service
agreements with ISPs
– edge routers mark packets as being “in” or “out” of profile
– core routers run RIO: RED with in/out
P(drop)
1.0
MaxP
Min in MaxinMaxoutMinout
AvgLen