15-441 Computer Networking

Lecture 18 – More TCP & Congestion

Control
Good Ideas So Far…

• Flow control
• Stop & wait
• Parallel stop & wait
• Sliding window (e.g., advertised windows)
• Loss recovery
• Timeouts
• Acknowledgement-driven recovery (selective repeat or cumulative
acknowledgement)
• Congestion control
• AIMD  fairness and efficiency

• How does TCP actually implement these?

Lecture 18: TCP Details 2

Outline

• The devilish details of TCP

• TCP connection setup and data transfer

• TCP reliability
• Be nice to your data

• TCP congestion avoidance

• Be nice to your routers

Lecture 18: TCP Details 3

Sequence Number Space

• Each byte in byte stream is numbered.

• 32 bit value
• Wraps around
• Initial values selected at start up time
• TCP breaks up the byte stream into packets.
• Packet size is limited to the Maximum Segment Size
• Each packet has a sequence number.
• Indicates where it fits in the byte stream

13450 14950 16050 17550

packet 8 packet 9 packet 10

Lecture 18: TCP Details 4

Establishing Connection:
Three-Way handshake
• Each side notifies other of
starting sequence number SYN: SeqC

it will use for sending

• Why not simply chose 0?
• Must avoid overlap with earlier ACK: SeqC+1
incarnation SYN: SeqS
• Security issues
• Each side acknowledges ACK: SeqS+1
other’s sequence number
• SYN-ACK: Acknowledge
sequence number + 1
• Can combine second SYN Client Server
with first ACK
Lecture 18: TCP Details 5
TCP Connection Setup Example

09:23:33.042318 IP 128.2.222.198.3123 > 192.216.219.96.80: S

4019802004:4019802004(0) win 65535 <mss 1260,nop,nop,sackOK> (DF)

09:23:33.118329 IP 192.216.219.96.80 > 128.2.222.198.3123: S

3428951569:3428951569(0) ack 4019802005 win 5840 <mss
1460,nop,nop,sackOK> (DF)

09:23:33.118405 IP 128.2.222.198.3123 > 192.216.219.96.80: . ack

3428951570 win 65535 (DF)

• Client SYN
• SeqC: Seq. #4019802004, window 65535, max. seg. 1260
• Server SYN-ACK+SYN
• Receive: #4019802005 (= SeqC+1)
• SeqS: Seq. #3428951569, window 5840, max. seg. 1460
• Client SYN-ACK
• Receive: #3428951570 (= SeqS+1)
Lecture 18: TCP Details 6
TCP State Diagram: Connection Setup
Client
CLOSED active OPEN
Server create TCB
passive OPEN CLOSE Snd SYN
create TCB delete TCB

LISTEN CLOSE
delete TCB

rcv SYN SEND

snd SYN ACK snd SYN
SYN rcv SYN SYN
RCVD snd ACK SENT
rcv ACK of SYN Rcv SYN, ACK
Snd ACK

CLOSE
Send FIN ESTAB

Lecture 18: TCP Details 7

Tearing Down Connection

• Either side can initiate tear

down A B
• Send FIN signal FIN, SeqA

• “I’m not going to send any more

data” ACK, SeqA+1

• Other side can continue Data

sending data ACK

• Half open connection

• Must continue to acknowledge FIN, SeqB
• Acknowledging FIN ACK, SeqB+1
• Acknowledge last sequence
number + 1

Lecture 18: TCP Details 8

TCP Connection Teardown Example
09:54:17.585396 IP 128.2.222.198.4474 > 128.2.210.194.6616: F
1489294581:1489294581(0) ack 1909787689 win 65434 (DF)

09:54:17.585732 IP 128.2.210.194.6616 > 128.2.222.198.4474: F

1909787689:1909787689(0) ack 1489294582 win 5840 (DF)

09:54:17.585764 IP 128.2.222.198.4474 > 128.2.210.194.6616: . ack

1909787690 win 65434 (DF)

• Session
• Echo client on 128.2.222.198, server on 128.2.210.194
• Client FIN
• SeqC: 1489294581
• Server ACK + FIN
• Ack: 1489294582 (= SeqC+1)
• SeqS: 1909787689
• Client ACK
• Ack: 1909787690 (= SeqS+1)
Lecture 18: TCP Details 9
State Diagram: Connection Tear-down

CLOSE
Active Close ESTAB
send FIN
CLOSE rcv FIN Passive Close
send FIN send ACK
FIN CLOSE
WAIT-1 WAIT
rcv FIN
ACK snd ACK CLOSE
rcv FIN+ACK snd FIN
FIN WAIT-2 snd ACK CLOSING LAST-ACK
rcv ACK of FIN rcv ACK of FIN

rcv FIN TIME WAIT CLOSED

Timeout=2msl
snd ACK delete TCB

Lecture 18: TCP Details 1

Outline

• TCP connection setup/data transfer

• TCP reliability

Lecture 18: TCP Details 1

Reliability Challenges

• Congestion related losses

• Variable packet delays
• What should the timeout be?
• Reordering of packets
• How to tell the difference between a delayed packet
and a lost one?

Lecture 18: TCP Details 1

TCP = Go-Back-N Variant

• Sliding window with cumulative acks

• Receiver can only return a single “ack” sequence number to the
sender.
• Acknowledges all bytes with a lower sequence number
• Starting point for retransmission
• Duplicate acks sent when out-of-order packet received
• But: sender only retransmits a single packet.
• Reason???
• Only one that it knows is lost
• Network is congested  shouldn’t overload it
• Error control is based on byte sequences, not packets.
• Retransmitted packet can be different from the original lost packet
– Why?

Lecture 18: TCP Details 1

• How to set timeout?
• Wait until sender knows it should have seen an ACK

• How long should this be?

Lecture 18: TCP Details 1

Round-trip Time Estimation

• Wait at least one RTT before retransmitting

• Importance of accurate RTT estimators:
• Low RTT estimate
• unneeded retransmissions
• High RTT estimate
• poor throughput
• RTT estimator must adapt to change in RTT
• But not too fast, or too slow!
• Spurious timeouts
• “Conservation of packets” principle – never more than a
window worth of packets in flight

Lecture 18: TCP Details 1

Original TCP Round-trip Estimator

• Round trip times

exponentially
averaged:
• New RTT = α (old RTT)
+ (1 - α) (new sample)
• Recommended value
for α: 0.8 - 0.9
• 0.875 for most TCP’s
• Retransmit timer set to (b * RTT), where b = 2
• Every time timer expires, RTO exponentially backed-off
• Not good at preventing spurious timeouts
• Why?

Lecture 18: TCP Details 1

RTT Sample Ambiguity

A B A B
Original trans Original trans
miss ion miss ion
X
RTO RTO
ACK
Sample retran
smiss Sample retran
io smiss
RTT n
RTT ion

ACK

• Karn’s RTT Estimator

• If a segment has been retransmitted:
• Don’t count RTT sample on ACKs for this segment
• Keep backed off time-out for next packet
• Reuse RTT estimate only after one successful transmission
Lecture 18: TCP Details 1
Jacobson’s Retransmission Timeout

• Key observation:
• At high loads round trip variance is high
• Solution:
• Base RTO on RTT and standard deviation
• RTO = RTT + 4 * rttvar
• new_rttvar = β * dev + (1- β) old_rttvar
• Dev = linear deviation
• Inappropriately named – actually smoothed linear
deviation

Lecture 18: TCP Details 1

Timestamp Extension

• Used to improve timeout mechanism by more

accurate measurement of RTT
• When sending a packet, insert current time into
option
• 4 bytes for time, 4 bytes for echo a received timestamp
• Receiver echoes timestamp in ACK
• Actually will echo whatever is in timestamp
• Removes retransmission ambiguity
• Can get RTT sample on any packet

Lecture 18: TCP Details 1

Timer Granularity

• Many TCP implementations set RTO in multiples

of 200,500,1000ms
• Why?
• Avoid spurious timeouts – RTTs can vary quickly due to
cross traffic
• Make timer interrupts efficient
• What happens for the first couple of packets?
• Pick a very conservative value (seconds)

Lecture 18: TCP Details 2

Fast Retransmit

• What are duplicate acks (dupacks)?

• Repeated acks for the same sequence
• When can duplicate acks occur?
• Loss
• Packet re-ordering
• Window update – advertisement of new flow control window
• Assume re-ordering is infrequent and not of large
magnitude
• Use receipt of 3 or more duplicate acks as indication of loss
• Don’t wait for timeout to retransmit packet

Lecture 18: TCP Details 2

 Fast Retransmit

Retransmission
X
Sequence No Duplicate Acks

Packets
Acks
Time

Lecture 18: TCP Details 2

 TCP (Reno variant)

X
X
X
Now what? - timeout
X
Sequence No

Packets
Acks
Time

Lecture 18: TCP Details 2

SACK

• Basic problem is that cumulative acks provide

little information
• Selective acknowledgement (SACK) essentially
adds a bitmask of packets received
• Implemented as a TCP option
• Encoded as a set of received byte ranges (max of 4
ranges/often max of 3)
• When to retransmit?
• Still need to deal with reordering  wait for out of order
by 3pkts

Lecture 18: TCP Details 2

 SACK

X
X
X
Now what? – send
X retransmissions as soon
Sequence No as detected

Packets
Acks
Time

Lecture 18: TCP Details 2

Performance Issues

• Timeout >> fast rexmit

• Need 3 dupacks/sacks

• Not great for small transfers

• Don’t have 3 packets outstanding

• What are real loss patterns like?

Lecture 18: TCP Details 2

Outline

• TCP connection setup/data transfer

• TCP reliability

• TCP congestion avoidance

10-30-2007 Lecture 18: TCP Details 26

Additive Increase/Decrease

• Both X1 and X2
increase/ decrease
by the same amount Fairness Line

over time T1

• Additive increase User 2’s

improves fairness and Allocation T0
x2
additive decrease
reduces fairness
Efficiency Line

User 1’s Allocation x1

10-30-2007 Lecture 18: TCP Details 27

Muliplicative Increase/Decrease

• Both X1 and X2
increase by the Fairness Line
same factor over T1

time
User 2’s
• Extension from Allocation
x2 T0
origin – constant
fairness
Efficiency Line

User 1’s Allocation x1

10-30-2007 Lecture 18: TCP Details 28

What is the Right Choice?

• Constraints limit
us to AIMD x1
Fairness Line

• Improves or
keeps fairness User 2’s
x0

constant at Allocation
x2 x2
each step
• AIMD moves
Efficiency Line
towards optimal
point User 1’s Allocation x1

10-30-2007 Lecture 18: TCP Details 29

TCP Congestion Control

• Changes to TCP motivated by ARPANET

congestion collapse
• Basic principles
• AIMD
• Packet conservation
• Reaching steady state quickly
• ACK clocking

10-30-2007 Lecture 18: TCP Details 30

AIMD

• Distributed, fair and efficient

• Packet loss is seen as sign of congestion and results in a
multiplicative rate decrease
• Factor of 2
• TCP periodically probes for available bandwidth by
increasing its rate

Rate

Time
10-30-2007 Lecture 18: TCP Details 31
Implementation Issue

• Operating system timers are very coarse – how to pace

packets out smoothly?
• Implemented using a congestion window that limits how
much data can be in the network.
• TCP also keeps track of how much data is in transit
• Data can only be sent when the amount of outstanding
data is less than the congestion window.
• The amount of outstanding data is increased on a “send” and
decreased on “ack”
• (last sent – last acked) < congestion window
• Window limited by both congestion and buffering
• Sender’s maximum window = Min (advertised window, cwnd)

10-30-2007 Lecture 18: TCP Details 32

Congestion Avoidance

• If loss occurs when cwnd = W

• Network can handle 0.5W ~ W segments
• Set cwnd to 0.5W (multiplicative decrease)
• Upon receiving ACK
• Increase cwnd by (1 packet)/cwnd
• What is 1 packet?  1 MSS worth of bytes
• After cwnd packets have passed by  approximately increase
of 1 MSS
• Implements AIMD

10-30-2007 Lecture 18: TCP Details 33

 Congestion Avoidance Sequence Plot

Sequence No

Packets
Acks
Time

10-30-2007 Lecture 18: TCP Details 34

Congestion Avoidance Behavior

Congestion
Window

Cut Grabbing
Time
Packet loss
+ retransmit Congestion back
Window Bandwidth
and Rate

10-30-2007 Lecture 18: TCP Details 35

Important Lessons

• TCP state diagram  setup/teardown

• TCP timeout calculation  how is RTT estimated

• Modern TCP loss recovery

• Why are timeouts bad?
• How to avoid them?  e.g. fast retransmit

10-30-2007 Lecture 18: TCP Details 36

Lecture 18: TCP Details 3