TCP - Part III
• Error Control
• Congestion Control
• Timer
© Jörg Liebeherr, 1998,1999
1
Error Control in TCP
• TCP implements a variation of the Go-back-N retransmission
scheme
• TCP maintains four (4) timers for each connection
• TCP couples error control and congestion control (I.e., it
assumes that errors are caused by congestion)
• TCP allows accelerated retransmissions (Fast Retransmit)
© Jörg Liebeherr, 1998,1999
2
1
Background on ARQ Error Control 1
• Two types of errors:
– Lost packets
– Damaged packets
• Most Error Control techniques are based on:
1. Error Detection Scheme (Parity checks, CRC).
2. Retransmission Scheme.
• Error control schemes that involve error detection and
retransmission of lost or corrupted frames are referred to as
Automatic Repeat Request (ARQ) error control.
© Jörg Liebeherr, 1998,1999
3
Background on ARQ Error Control 2
All retransmission schemes use all or a subset of the following
procedures:
Positive acknowledgments (ACK)
Negative acknowledgment (NACK)
All retransmission schemes (using ACK, NACK or both)
rely on the use of timers
The most common ARQ retransmission schemes are:
Stop-and-Wait ARQ
Go-Back-N ARQ
Selective Repeat ARQ
© Jörg Liebeherr, 1998,1999
4
2
Error Control in TCP
• TCP implements a variation of the Go-back-N retransmission
scheme
• TCP maintains four (4) timers for each connection
• TCP couples error control and congestion control (I.e., it
assumes that errors are caused by congestion)
• TCP allows accelerated retransmissions (Fast Retransmit)
© Jörg Liebeherr, 1998,1999
5
TCP Timers
• TCP maintains four (4) timers for each connection:
– Retransmission Timer:
• The timer is started during a transmission. A timeout causes a
retransmission
– Persist Timer
• Ensures that window size information is transmitted even if no data
is transmitted
– Keepalive Timer
• Detects crashes on the other end of the connection
– 2MSL Timer
• Measures the time that a connection has been in the TIME_WAIT
state
© Jörg Liebeherr, 1998,1999
6
3
TCP Retransmission Timer
• Retransmission Timer:
– The setting of the retransmission timer is crucial for
efficiency
– Timeout value too small -> results in unnecessary
retransmissions
– Timeout value too large -> long waiting time before a
retransmission can be issued
– A problem is that the delays in the network are not fixed
– Therefore, the retransmission timers must be adaptive
© Jörg Liebeherr, 1998,1999
7
Measuring TCP Retransmission Timers
ftp session
from aida
to rigoletto
aida.poly.edu
rigoletto.poly.edu
•Transfer file from aida to rigoletto
• Unplug Ethernet cable in the middle of file transfer
© Jörg Liebeherr, 1998,1999
8
4
tcpdump Trace
10:42:01.704681 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:42:01.705603 aida.40001 > rigoletto.ftp-data: . 162649:164109(1460) ack 1 win 17520
10:42:01.706753 aida.40001 > rigoletto.ftp-data: . 164109:165569(1460) ack 1 win 17520
10:42:02.741764 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:42:05.741788 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:42:11.741828 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:42:23.741951 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:42:47.742176 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:43:35.742587 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:44:39.743140 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:45:43.743702 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:46:47.744271 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:47:51.752138 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:48:55.745547 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:49:59.746123 aida.40001 > rigoletto.ftp-data: . 161189:162649(1460) ack 1 win 17520
10:51:03.745839 aida.40001 > rigoletto.ftp-data: R 165569:165569(0) ack 1 win 17520
© Jörg Liebeherr, 1998,1999
9
Interpreting the Measurements
© Jörg Liebeherr, 1998,1999
600
500
400
300
200
12
10
8
6
4
0
2
100
0
Seconds
• The interval between
retransmission attempts in
seconds is:
1.03, 3, 6, 12, 24, 48, 64,
64, 64, 64, 64, 64, 64.
• Time between retransmissions is doubled each
time (Exponential Backoff
Algorithm)
• Timer is not increased
beyond 64 seconds
• TCP gives up after 13th
attempt and 9 minutes.
Transmission Attempts
10
5
Round-Trip Time Measurements
• The retransmission mechanism of TCP is adaptive
• The retransmission timers are set based on round-trip time
(RTT) measurements that TCP performs
Segment 1
RTT #1
Segment 2
Segment 3
RTT #2
ACK for Se
gment 2 +
Segment
5
RTT #3
The RTT is based on time difference
between segment transmission and
ACK
But:
TCP does not ACK each
segment
Each connection has only one
timer
ent 1
ACK for Segm
ACK for Se
ACK for Se
3
Segmen
t4
gment 4
gment 5
© Jörg Liebeherr, 1998,1999
11
Round-Trip Time Measurements
• Retransmission timer is set to a Retransmission Timeout
(RTO) value.
• RTO is calculated based on the RTT measurements.
• The RTT measurements are smoothed by the following
estimators srtt and rttvar:
srttn+1 = α RTT + (1- α ) srttn
rttvarn+1 = β ( | RTT - srttn+1 | ) + (1- β ) rttvarn
RTOn+1 = srttn+1 + 4 rttvarn+1
• The gains are set to α =1/4 and β =1/8
• srtt0 = 0 sec, rttvar0 = 3 sec, Also: RTO1 = srtt1 + 2 rttvar1
© Jörg Liebeherr, 1998,1999
12
6
Karn’s Algorithm
• If an ACK for a retransmitted
segment is received, the sender
cannot tell if the ACK belongs to
the original or the retransmission.
segme
nt
retrans
mission
of segm
ent
Timeout !
RTT ?
RTT ?
ACK
Karn’s Algorithm:
Don’t update srtt on any segments that have been retransmitted.
Each time when TCP retransmits, it sets:
RTOn+1 = max ( 2 RTOn, 64)
(exponential backoff)
© Jörg Liebeherr, 1998,1999
13
RTO Calculation: Example
• At t1: RTO = srtt + 2 rttvar = 6 sec
• At t2: RTO= 2 * (srtt + 4rttvar) = 24
sec
(exponential backoff)
• At t4: RTO is not updated
.
.
Seg
men
Seg
t4
me
nt 5
Seg
me
nt 6
.
Segm
ent 2
Segm
ent 3
Segmen
t1
ACK
SYN
for
ACK t 4
men
Seg
for
t3
ACK
men
Seg
ent 2
r Segm
ACK fo
ent 1
r Segm
ACK fo
ACK
SYN +
SYN
(Due to Karn’s algorithm)
Timeout !
RTT #2
RTT #1
t1
© Jörg Liebeherr, 1998,1999
t2 t3
t4
t5
t6
RTT #3
t7
t8
t9
14
7
Congestion Avoidance
• Most often, a packet loss in a network is due to an overflow at
a congested router (rather than due to a transmission error)
• A sender can detect lost packets through a:
• Timeout of a retransmission timer
• Receipt of a duplicate ACK
• TCP assumes that a packet loss is caused by congestion and
reduces the size of the sending window
• The algorithm that reduces and then reopens the sending
window is called Congestion Avoidance
© Jörg Liebeherr, 1998,1999
15
Slow Start / Congestion Avoidance
• Implemented with two variables:
• Congestion window (cwnd)
• Slow start threshhold (ssthr)
• Initialization:
cwnd= MSS bytes
ssthr = <advertised window at setup>
• The window size at the sender is set as follows:
Allowed Window = MIN (advertised window, cwnd)
© Jörg Liebeherr, 1998,1999
16
8
Slow Start / Congestion Avoidance
• Here we give a more accurate version than in our earlier
discussion of Slow Start:
If cwnd <= ssthresh then
/* Slow Start Phase */
Each time a segment is acknowledged:
cwnd = cwnd + MSS
else /* cwnd > ssthresh */
/* Congestion Avoidance Phase */
Each time a segment is acknowledged:
cwnd = cwnd + MSS * MSS / cwnd + segsize / 8
endif
© Jörg Liebeherr, 1998,1999
17
Slow Start / Congestion Avoidance
• Each time when congestion occurs (timeout or receipt of
duplicate ACK),
– cwnd is reset to one:
cwnd = 1
– ssthresh is set to half the current size of the congestion
window:
ssthressh = cwnd / 2
© Jörg Liebeherr, 1998,1999
18
9
Slow Start / Congestion Avoidance
• A typical plot of cwnd for a TCP connection (segsize = 1500
bytes) :
© Jörg Liebeherr, 1998,1999
19
Fast Retransmit
• After three identical ACKs
are received by the sender,
it transmits a single segment
without waiting for a timeout
to expire.
Data (100:200)
ACK 100
ACK 100
ACK 100
Data (100:200
)
If 3rd duplicate ACK is received:
ssthresh = cwnd / 2
cwnd = ssthresh + 3 segsize
retransmit segment
For each additional duplicate:
cwnd = cwnd + segsize
and transmit a segment
When an ACK arrives that acknowledges new data:
cwnd = ssthresh
© Jörg Liebeherr, 1998,1999
20
10
Repacketization
• When TCP does a retransmission, it can send the missing
data in differently sized segments
Data (1:100)
ACK 100
Data (100:200
)
lost
Data (100
:200)
ACK 200
© Jörg Liebeherr, 1998,1999
21
TCP Persist Timer
• Assume the window size goes down to zero and the ACK that
opens the window gets lost
If ACK (see figure) is lost,
both sides are blocked.
Receiver
Buffer
0
4K
2K SeqNo=0
Persist Timer:
2K
Win=2048
AckNo=2048
2K SeqNo=20
48
Sender blocked
Forces that the sender
periodically queries the receiver
about its window size (window
probes)
4K
Win=0
AckNo=4096
3K
ACK is
lost
© Jörg Liebeherr, 1998,1999
Win=1024
AckNo=4096
22
11
TCP Persist Timer
• The persist timer is started by the sender when the sliding
window is zero
Persist timer uses exponential backoff (initial value is 1.5 seconds)
rounded to the range [5 sec, 60sec]
So the time interval between timeouts are at:
5, 5, 6, 12, 24, 48, 60, 60, ...
The window probe packet contains one byte of data (TCP can do
this even if the window size is zero)
© Jörg Liebeherr, 1998,1999
23
TCP Persist Timer
Timeout
(5 sec)
Win=0
AckNo=4096
Probe
1 byte SeqN
o=4096
Timeout
(5 sec)
Win=0
AckNo=4096
Probe
1 byte SeqN
o=4096
Timeout
(6 sec)
Win=0
AckNo=4096
Probe
1 byte SeqNo=
4096
Win=1024
AckNo=4096
1 KB
© Jörg Liebeherr, 1998,1999
SeqNo=5120
24
12
TCP Keepalive Timer
• When a TCP connection has been idle for a long time, a
Keepalive timer reminds a station to check if the other side is
still there.
• A probe packet is sent if the connection has been idle for 2
hours
• Assume a probe has been sent from A to B:
(1) B is up and running:
B responds with an ACK
(2) B has crashed and is down:
A will send 10 more probes, each 75
seconds apart. If A does not get a
response, it will close the connection
(3) B has rebooted:
B will send a RST segment
(4) B is up, but unreachable:
Looks to A the same as (2)
© Jörg Liebeherr, 1998,1999
25
13