Professional Documents
Culture Documents
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
provide logical communication between app processes running on different hosts transport protocols run in end systems send side: breaks app messages into segments, passes to network layer rcv side: reassembles segments into messages, passes to app layer more than one transport protocol available to apps
household analogy:
12 kids in Anns house sending letters to 12 kids in Bills house: hosts = houses processes = kids app messages = letters in envelopes transport protocol = Ann and Bill who demux to inhouse siblings network-layer protocol = postal service
application transport network data link physical network data link physical
network data link physical network data link physical network data link physical network data link physical application transport network data link physical
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
Multiplexing/demultiplexing
multiplexing at sender: handle data from multiple sockets, add transport header (later used for demultiplexing)
application application
demultiplexing at receiver: use header info to deliver received segments to correct socket
P2
P1
application
P3
transport network link physical
transport network
P4
transport network link physical
socket process
link
physical
32 bits
source port # dest port #
host uses IP addresses & port numbers to direct segment to appropriate socket
Connectionless demultiplexing
recall: when creating datagram to send into UDP socket, must specify
destination IP address destination port #
IP datagrams with same dest. port #, but different source IP addresses and/or source port numbers will be directed to same socket at dest
Transport Layer 3-10
P1
transport
P3
transport network link physical source port: 6428 dest port: 9157 source port: ? dest port: ? network link physical
P4
transport network link physical
Connection-oriented demux
demux: receiver uses all four values to direct segment to appropriate socket
P4
P5
transport
P6
application
P3
transport network link physical network link physical
P2
P3
server: IP address B
source IP,port: B,80 dest IP,port: A,9157 source IP,port: A,9157 dest IP, port: B,80
host: IP address A
host: IP address C
three segments, all destined to IP address: B, dest port: 80 are demultiplexed to different sockets
threaded server
application
P4
transport
P3
transport network link physical network link physical
P2
P3
server: IP address B
source IP,port: B,80 dest IP,port: A,9157 source IP,port: A,9157 dest IP, port: B,80
host: IP address A
host: IP address C
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
no frills, bare bones Internet transport protocol best effort service, UDP segments may be: lost delivered out-of-order to app connectionless: no handshaking between UDP sender, receiver each UDP segment handled independently of others
UDP use:
streaming multimedia apps (loss tolerant, rate sensitive) DNS SNMP
no connection establishment (which can add delay) simple: no connection state at sender, receiver small header size no congestion control: UDP can blast away as fast as desired
Transport Layer 3-17
UDP checksum
Goal: detect errors (e.g., flipped bits) in transmitted segment
sender:
receiver:
treat segment contents, including header fields, as sequence of 16-bit integers checksum: addition (ones complement sum) of segment contents sender puts checksum value into UDP checksum field
compute checksum of received segment check if computed checksum equals checksum field value: NO - error detected YES - no error detected. But maybe errors nonetheless? More later .
Transport Layer 3-18
Note: when adding numbers, a carryout from the most significant bit needs to be added to the result
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 3-21
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 3-22
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 3-23
send side
receive side
state: when in this state next state uniquely determined by next event
state 1
state 2
sender
receiver
Transport Layer 3-26
receiver
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
udt_send(NAK)
sender
Wait for call from below rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) extract(rcvpkt,data) deliver_data(data) udt_send(ACK)
Transport Layer 3-29
Wait for call from below rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) extract(rcvpkt,data) deliver_data(data) udt_send(ACK)
Transport Layer 3-30
Wait for call from below rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) extract(rcvpkt,data) deliver_data(data) udt_send(ACK)
Transport Layer 3-31
handling duplicates:
sender retransmits current pkt if ACK/NAK sender doesnt know corrupted what happened at sender adds sequence receiver! number to each pkt can't just retransmit: receiver discards possible duplicate (doesnt deliver up) stop and wait duplicate pkt
L
Wait for ACK or NAK 1
rdt_rcv(rcvpkt) && (corrupt(rcvpkt) sndpkt = make_pkt(NAK, chksum) udt_send(sndpkt) rdt_rcv(rcvpkt) && not corrupt(rcvpkt) && has_seq1(rcvpkt) sndpkt = make_pkt(ACK, chksum) udt_send(sndpkt)
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) && has_seq1(rcvpkt) extract(rcvpkt,data) deliver_data(data) sndpkt = make_pkt(ACK, chksum) udt_send(sndpkt)
rdt2.1: discussion
sender: seq # added to pkt two seq. #s (0,1) will suffice. Why? must check if received ACK/NAK corrupted twice as many states
state must remember whether expected pkt should have seq # of 0 or 1
note: receiver can not know if its last ACK/NAK received OK at sender
Transport Layer 3-35
same functionality as rdt2.1, using ACKs only instead of NAK, receiver sends ACK for last pkt received OK
receiver must explicitly include seq # of pkt being ACKed
duplicate ACK at sender results in same action as NAK: retransmit current pkt
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) && has_seq1(rcvpkt) extract(rcvpkt,data) deliver_data(data) sndpkt = make_pkt(ACK1, chksum) udt_send(sndpkt)
retransmits if no ACK received in this time if pkt (or ACK) just delayed (not lost): retransmission will be duplicate, but seq. #s already handles this receiver must specify seq # of pkt being ACKed requires countdown timer
Transport Layer 3-38
rdt3.0 sender
rdt_send(data) sndpkt = make_pkt(0, data, checksum) udt_send(sndpkt) start_timer Wait for call 0from above Wait for ACK0 rdt_rcv(rcvpkt) && ( corrupt(rcvpkt) || isACK(rcvpkt,1) ) rdt_rcv(rcvpkt)
L
timeout udt_send(sndpkt) start_timer rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) && isACK(rcvpkt,0) stop_timer
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) && isACK(rcvpkt,1) stop_timer Wait for ACK1 rdt_send(data)
rdt_rcv(rcvpkt)
rdt3.0 in action
sender
send pkt0 rcv ack0 send pkt1 rcv ack1 send pkt0
pkt0 ack0 pkt1 ack1 pkt0 ack0
receiver
rcv pkt0 send ack0 rcv pkt1 send ack1 rcv pkt0 send ack0
sender
send pkt0 rcv ack0 send pkt1
pkt0 ack0 pkt1
receiver
rcv pkt0 send ack0
loss
resend pkt1
timeout
(a) no loss
rdt3.0 in action
sender
send pkt0 rcv ack0 send pkt1
pkt0 ack0 pkt1
sender
send pkt0 rcv ack0 send pkt1
pkt0 ack0 pkt1
receiver
rcv pkt0 send ack0 rcv pkt1 send ack1
receiver
rcv pkt0 send ack0 rcv pkt1 send ack1
loss
ack1
ack1
resend pkt1
timeout
pkt1
ack1
pkt0 ack0
(detect duplicate)
rcv pkt1
send ack1
resend pkt1 rcv ack1 send pkt0 rcv ack1 send pkt0
timeout
pkt1
(detect duplicate)
rcv pkt1
(detect duplicate)
send ack1 rcv pkt0 send ack0 rcv pkt0 send ack0
Performance of rdt3.0
rdt3.0 is correct, but performance stinks e.g.: 1 Gbps link, 15 ms prop. delay, 8000 bit packet:
Dtrans = R = 109 bits/sec
8000 bits
= 8 microsecs
L/R
RTT + L / R
.008
30.008
= 0.00027
if RTT=30 msec, 1KB pkt every 30 msec: 33kB/sec thruput over 1 Gbps link
RTT
sender =
L/R RTT + L / R
.008
30.008
= 0.00027
Pipelined protocols
pipelining: sender allows multiple, in-flight, yet-to-be-acknowledged pkts
range of sequence numbers must be increased buffering at sender and/or receiver
RTT
sender =
3L / R RTT + L / R
.0024
30.008
= 0.00081
Selective Repeat: sender can have up to N unacked packets in pipeline rcvr sends individual ack for each packet
Go-Back-N: sender
ACK(n): ACKs all pkts up to, including seq # n cumulative ACK may receive duplicate ACKs (see receiver) timer for oldest in-flight pkt timeout(n): retransmit packet n and all higher seq # pkts in window
Transport Layer 3-47
L
base=1 nextseqnum=1
Wait
rdt_rcv(rcvpkt) && corrupt(rcvpkt)
rdt_rcv(rcvpkt) && notcorrupt(rcvpkt) base = getacknum(rcvpkt)+1 If (base == nextseqnum) stop_timer else start_timer
ACK-only: always send ACK for correctly-received pkt with highest in-order seq #
may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt:
discard (dont buffer): no receiver buffering! re-ACK pkt with highest in-order seq #
Transport Layer 3-49
GBN in action
sender window (N=4)
012345678
012345678 012345678 012345678
sender
send pkt0 send pkt1 send pkt2 send pkt3 (wait)
receiver
receive pkt0, send ack0 receive pkt1, send ack1 receive pkt3, discard, (re)send ack1 receive pkt4, discard, (re)send ack1 receive pkt5, discard, (re)send ack1
Xloss
012345678 012345678
pkt 2 timeout
012345678 012345678 012345678 012345678
Selective repeat
sender window
N consecutive seq #s limits seq #s of sent, unACKed pkts
Selective repeat
sender data from above:
timeout(n):
resend pkt n, restart timer ACK(n) in
[sendbase,sendbase+N]:
send ACK(n) out-of-order: buffer in-order: deliver (also deliver buffered, inorder pkts), advance window to next not-yetreceived pkt ACK(n)
mark pkt n as received if n smallest unACKed pkt, advance window base to next unACKed seq #
pkt n in [rcvbase-N,rcvbase-1]
otherwise:
ignore
Transport Layer 3-53
sender
send pkt0 send pkt1 send pkt2 send pkt3 (wait)
receiver
receive pkt0, send ack0 receive pkt1, send ack1 receive pkt3, buffer, send ack3 receive pkt4, buffer, send ack4 receive pkt5, buffer, send ack5
012345678
012345678 012345678
Xloss
012345678 012345678
pkt 2 timeout
012345678 012345678 012345678 012345678
send pkt2
rcv pkt2; deliver pkt2, pkt3, pkt4, pkt5; send ack2
seq #s: 0, 1, 2, 3 window size=3 receiver sees no difference in two scenarios! duplicate data accepted as new in (b)
X
will accept packet with seq number 0
(a) no problem
receiver cant see sender side. receiver behavior identical in both cases! somethings (very) wrong!
0123012
0123012
0123012
Q: what relationship between seq # size and window size to avoid problem in (b)?
pkt0
(b) oops!
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
TCP: Overview
2581
point-to-point:
one sender, one receiver
connection-oriented:
handshaking (exchange of control msgs) inits sender, receiver state before data exchange
pipelined:
TCP congestion and flow control set window size
flow controlled:
sender will not Transport Layer 3-57 overwhelm receiver
source port #
dest port #
sequence number
acknowledgement number
head not UAP R S F len used
checksum
sequence numbers: byte stream number of first byte in segments data acknowledgements: seq # of next byte expected from other side cumulative ACK Q: how receiver handles out-of-order segments A: TCP spec doesnt say, - up to implementor
source port #
dest port #
window size
User types C
Seq=43, ACK=80
too short: premature timeout, unnecessary retransmissions too long: slow reaction to segment loss
SampleRTT: measured time from segment transmission until ACK receipt ignore retransmissions SampleRTT will vary, want estimated RTT smoother average several recent measurements, not just current SampleRTT
Transport Layer 3-61
exponential weighted moving average influence of past sample decreases exponentially fast typical value: = 0.125
RTT: gaia.cs.umass.edu to fantasia.eurecom.fr
350
RTT (milliseconds)
300
RTT (milliseconds)
250
200
sampleRTT
150
EstimatedRTT
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
100
time (seconds)
SampleRTT
time (seconnds)
Estimated RTT
safety margin
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
update what is known to be ACKed start timer if there are still unacked segments
Transport Layer 3-66
L
NextSeqNum = InitialSeqNum SendBase = InitialSeqNum
timeout
retransmit not-yet-acked segment with smallest seq. # start timer
SendBase=92 Seq=92, 8 bytes of data timeout ACK=100 timeout Seq=92, 8 bytes of data Seq=100, 20 bytes of data
ACK=100 ACK=120 Seq=92, 8 bytes of data SendBase=100 ACK=100 SendBase=120 ACK=120 SendBase=120 Seq=92, 8 bytes of data
premature timeout
Transport Layer 3-68
ACK=100
ACK=120
cumulative ACK
Transport Layer 3-69
immediate send ACK, provided that segment starts at lower end of gap
Transport Layer 3-70
if sender receives 3 ACKs for same data (triple duplicate ACKs), resend unacked segment with smallest seq #
likely that unacked segment lost, so dont wait for timeout
Host B
X
ACK=100 timeout ACK=100 ACK=100 ACK=100 Seq=100, 20 bytes of data
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
TCP code
receiver controls sender, so sender wont overflow receivers buffer by transmitting too much, too fast
flow control
IP code
from sender
receiver advertises free buffer space by including rwnd value in TCP header of receiver-tosender segments
RcvBuffer size set via socket options (typical default is 4096 bytes) many operating systems autoadjust RcvBuffer
to application process
RcvBuffer rwnd
sender limits amount of unacked (in-flight) data to receivers rwnd value guarantees receive buffer will not overflow
receiver-side buffering
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
Connection Management
before exchanging data, sender/receiver handshake:
agree to establish connection (each knowing the other willing to establish connection) agree on connection parameters
application
connection state: ESTAB connection variables: seq # client-to-server server-to-client rcvBuffer size at server,client
application
connection state: ESTAB connection Variables: seq # client-to-server server-to-client rcvBuffer size at server,client
network
network
choose x ESTAB
req_conn(x)
acc_conn(x) ESTAB
variable delays retransmitted messages (e.g. req_conn(x)) due to message loss message reordering can't see other side
req_conn(x)
ESTAB
acc_conn(x) retransmit req_conn(x) ESTAB req_conn(x)
connection x completes
retransmit data(x+1)
server forgets x ESTAB client terminates
connection x completes
accept data(x+1)
client terminates
req_conn(x) data(x+1)
server forgets x
ESTAB accept data(x+1)
server state
LISTEN
SYNSENT
SYNbit=1, Seq=x
choose init seq num, y send TCP SYNACK SYN RCVD msg, acking SYN
received SYNACK(x) indicates server is live; ESTAB send ACK for SYNACK; this segment may contain client-to-server data
ACKbit=1, ACKnum=y+1
received ACK(y) indicates client is live
ESTAB
L
SYN(x)
SYNACK(seq=y,ACKnum=x+1) create new socket for communication back to client
listen
SYN(seq=x)
SYN sent
SYNACK(seq=y,ACKnum=x+1)
ACK(ACKnum=y+1)
ACK(ACKnum=y+1)
L
Transport Layer 3-81
server state
ESTAB
can no longer send but can receive data wait for server close
FIN_WAIT_1
FIN_WAIT_2
LAST_ACK
can no longer send data
CLOSED
Transport Layer 3-83
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
informally: too many sources sending too much data too fast for network to handle different from flow control! manifestations: lost packets (buffer overflow at routers) long delays (queueing in router buffers) a top-10 problem!
two senders, two receivers one router, infinite buffers output link capacity: R no retransmission
throughput:
lout
Host B
R/2
delay
lout
lout
Host A
Host B
lout lin
R/2
copy
lout
Host A
Host B
lost, dropped at router due to full buffers sender only resends if packet known to be lost
copy
lout
Host A
no buffer space!
Host B
Transport Layer 3-89
R/2 when sending at R/2, some packets are retransmissions but asymptotic goodput is still R/2 (why?)
lost, dropped at router due to full buffers sender only resends if packet known to be lost
lout
lin
R/2
lout
Host B
Transport Layer 3-90
packets can be lost, dropped at router due to full buffers sender times out prematurely, sending two copies, both of which are delivered lin
timeout copy
R/2 when sending at R/2, some packets are retransmissions including duplicated that are delivered!
lout
lin
R/2
l'in
lout
Host B
Transport Layer 3-91
packets can be lost, dropped at router due to full buffers sender times out prematurely, sending two copies, both of which are delivered
R/2 when sending at R/2, some packets are retransmissions including duplicated that are delivered!
lout
lin
R/2
costs of congestion:
more work (retrans) for given goodput unneeded retransmissions: link carries multiple copies of pkt decreasing goodput
lin increase ? A: as red lin increases, all arriving blue pkts at upper queue are dropped, blue throughput l g 0 out
Host B
Host D Host C
lout
lin
C/2
another cost of congestion: when packet dropped, any upstream transmission capacity used for that packet was wasted!
Transport Layer 3-94
no explicit feedback from network congestion inferred from end-system observed loss, delay approach taken by TCP
routers provide feedback to end systems single bit indicating congestion (SNA, DECbit, TCP/IP ECN, ATM) explicit rate for sender to send at
Transport Layer 3-95
elastic service if senders path underloaded: sender should use available bandwidth if senders path congested: sender throttled to minimum guaranteed rate
sent by sender, interspersed with data cells bits in RM cell set by switches (networkassisted) NI bit: no increase in rate (mild congestion) CI bit: congestion indication RM cells returned to sender by receiver, with bits intact
Transport Layer 3-96
Chapter 3 outline
3.1 transport-layer services 3.2 multiplexing and demultiplexing 3.3 connectionless transport: UDP 3.4 principles of reliable data transfer 3.5 connection-oriented transport: TCP
segment structure reliable data transfer flow control connection management
approach: sender increases transmission rate (window size), probing for usable bandwidth, until loss occurs additive increase: increase cwnd by 1 MSS every RTT until loss detected multiplicative decrease: cut cwnd in half after additively increase window size loss . until loss occurs (then cut window in half)
time
Transport Layer 3-99
TCP sending rate: roughly: send cwnd bytes, wait RTT for ACKS, then send more bytes
rate
~ ~
cwnd
RTT
bytes/sec
< cwnd
when connection begins, increase rate exponentially until first loss event:
initially cwnd = 1 MSS double cwnd every RTT done by incrementing cwnd for every ACK received
Host A
Host B
RTT
time
or 3 duplicate acks)
Transport Layer 3-102
Implementation:
variable ssthresh on loss event, ssthresh is set to 1/2 of cwnd just before loss event
Transport Layer 3-103
New ACK!
new ACK cwnd = cwnd+MSS dupACKcount = 0 transmit new segment(s), as allowed cwnd > ssthresh L timeout ssthresh = cwnd/2 cwnd = 1 MSS dupACKcount = 0 retransmit missing segment
new ACK cwnd = cwnd + MSS (MSS/cwnd) dupACKcount = 0 transmit new segment(s), as allowed
New ACK!
slow start
congestion avoidance
duplicate ACK dupACKcount++
New ACK!
New ACK cwnd = ssthresh dupACKcount = 0 dupACKcount == 3 ssthresh= cwnd/2 cwnd = ssthresh + 3 retransmit missing segment
fast recovery
duplicate ACK
TCP throughput
4 RTT
bytes/sec
W/2
example: 1500 byte segments, 100ms RTT, want 10 Gbps throughput requires W = 83,333 in-flight segments throughput in terms of segment loss probability, L [Mathis 1997]:
. MSS 1.22 TCP throughput = RTT L
to achieve 10 Gbps throughput, need a loss rate of L = 210-10 a very small loss rate!
TCP Fairness
fairness goal: if K TCP sessions share same bottleneck link of bandwidth R, each should have average rate of R/K
TCP connection 1
TCP connection 2
additive increase gives slope of 1, as throughout increases multiplicative decrease decreases throughput equal bandwidth share R proportionally
loss: decrease window by factor of 2 congestion avoidance: additive increase loss: decrease window by factor of 2 congestion avoidance: additive increase
Connection 1 throughput
R
Transport Layer 3-108
Fairness (more)
Fairness and UDP multimedia apps often do not use TCP
do not want rate throttled by congestion control
Fairness, parallel TCP connections application can open multiple parallel connections between two hosts web browsers do this e.g., link of rate R with 9 existing connections:
new app asks for 1 TCP, gets rate R/10 new app asks for 11 TCPs, gets R/2
Transport Layer 3-109
Chapter 3: summary
principles behind transport layer services: multiplexing, demultiplexing reliable data transfer flow control congestion control instantiation, implementation in the Internet
UDP TCP
next: leaving the network edge (application, transport layers) into the network core