Why Internetworking?
To build a “network of networks” or internet
operating over multiple, coexisting, different network
technologies
providing ubiquitous connectivity through IP packet transfer achieving huge economies of scale
H
H
Net 3 Net 5
Net 1 Net 5
G
G
G
G
Net 5 Net 5
G
H
G
Net 2 Net 5
Net 4 Net 5
H
Why Internetworking?
To provide universal communication services independent of underlying network technologies
providing common interface to user applications
Reliable Stream Service
User Datagram Service
H
H
Net 3 Net 5
Net 1 Net 5
G
G
G
G
Net 5 Net 5
G
H
G
Net 2 Net 5
Net 4 Net 5
H
Why Internetworking?
To provide distributed applications
Any application designed to operate based on Internet communication services immediately operates across the entire Internet
Rapid deployment of new applications Email, WWW, Peer-to-peer
Applications independent of network technology New networks can be introduced below Old network technologies can be retired
Internet Protocol Approach
IP packets transfer information across Internet Host A IP → router→ router…→ router→ Host B IP IP layer in each router determines next hop (router)
Network interfaces transfer IP packets across networks
Host A
Router
Router
Host B
Internet Layer
Network Interface
Transport Layer
Internet Layer
Network Interface
Transport Layer
Internet Layer
Network Interface
Internet Layer
Network Interface
Net 1 Net 5
Router
Internet Layer
Network Interface
Net 4 Net 5
Net 3 Net 5
Net 2 Net 5
TCP/IP Protocol Suite Distributed
HTTP
SMTP
DNS
RTP
Reliable stream service
Best-effort connectionless packet transfer
applications
User datagram service
(ICMP, ARP)
TCP
UDP
IP
Network Interface 3
Network Interface 1
Network Interface 2
Diverse network technologies
Internet Names & Addresses
Internet Names
Each host has a unique name
Independent of physical
location
Facilitate memorization by humans
Domain Name
Organization under single
administrative unit Host Name
Name given to host computer
User Name
Name assigned to user
Internet Addresses
leongarcia@comm.utoronto.ca
DNS resolves IP name
(intj = jth octet)
Each host has globally unique logical 32 bit IP address
Separate address for each physical connection to a network
Routing decision is done based on destination IP address
IP address has two parts:
netid and hostid
netid unique
netid facilitates routing Dotted Decimal Notation:
int1.int2.int3.int4
to IP address 128.100.10.13
Physical Addresses
LANs (and other networks) assign physical addresses to the physical attachment to the network
The network uses its own address to transfer packets or frames to the appropriate destination
IP address needs to be resolved to physical address at each IP network interface
Example: Ethernet uses 48-bit addresses
Each Ethernet network interface card (NIC) has globally
unique Medium Access Control (MAC) or physical address
First 24 bits identify NIC manufacturer; second 24 bits are serial number
00:90:27:96:68:07 12 hex numbers Intel
Encapsulation
TCP Header contains source & destination port numbers
HTTP Request
IP Header contains source and destination IP addresses; transport protocol type
TCP header
HTTP Request
Ethernet Header contains source & destination MAC addresses;
network protocol type
IP header
TCP header
HTTP Request
Ethernet header
IP header
TCP header
HTTP Request
FCS
Internet Protocol
Provides best effort, connectionless packet delivery motivated by need to keep routers simple and by
adaptibility to failure of network elements
packets may be lost, out of order, or even duplicated
higher layer protocols must deal with these, if necessary
RFCs 791, 950, 919, 922, and 2474.
Internet Control Message Protocol (ICMP), RFC 792
Internet Group Management Protocol (IGMP), RFC 1112
IP Packet Header
048 161924 31
Version
IHL
Type of Service
Total Length
Identification
Flags
Fragment Offset
Time to Live
Protocol
Header Checksum
Source IP Address
Destination IP Address
Options
Padding
Minimum 20 bytes
Up to 40 bytes in options fields
IP Packet Header
048 161924 31
Version
IHL
Type of Service
Total Length
Identification
Flags
Fragment Offset
Time to Live
Protocol
Header Checksum
Source IP Address
Destination IP Address
Options
Padding
Version: current IP version is 4.
Internet header length (IHL): length of the header in 32-bit words.
Type of service (TOS): traditionally priority of packet at each router. Recent Differentiated Services redefines TOS field to include other services besides best effort.
IP Packet Header
048 161924 31
Version
IHL
Type of Service
Total Length
Identification
Flags
Fragment Offset
Time to Live
Protocol
Header Checksum
Source IP Address
Destination IP Address
Options
Padding
Total length: number of bytes of the IP packet including header and data, maximum length is 65535 bytes.
Identification, Flags, and Fragment Offset: used for fragmentation and reassembly (More on this shortly).
Fragmentation and Reassembly
• Identification identifies a particular packet
• Flags = (unused, don’t fragment/DF, more fragment/MF)
• Fragment offset identifies the location of a fragment within a packet
Fragment at source
Source
Router
Reassemble at destination
Destination
Fragment at router
IP
IP
Network
Network
Example: Fragmenting a Packet
A packet is to be forwarded to a network with MTU of 576 bytes. The packet has an IP header of 20 bytes and a data part of 1484 bytes. and of each fragment.
Maximum data length per fragment = 576 – 20 = 556 bytes.
We set maximum data length to 552 bytes to get multiple of 8.
Total Length
Id
MF
Fragment Offset
Original packet
1504
x
0
0
Fragment 1
572
x
1
0
Fragment 2
572
x
1
69
Fragment 3
400
x
0
138
IP Packet Header
048 161924 31
Version
IHL
Type of Service
Total Length
Identification
Flags
Fragment Offset
Time to Live
Protocol
Header Checksum
Source IP Address
Destination IP Address
Options
Padding
Time to live (TTL): number of hops packet is allowed to traverse in the network. • Each router along the path to the destination decrements this value by one.
• If the value reaches zero before the packet reaches the destination, the router discards the packet and sends an error message back to the source.
IP Packet Header
048 161924 31
Version
IHL
Type of Service
Total Length
Identification
Flags
Fragment Offset
Time to Live
Protocol
Header Checksum
Source IP Address
Destination IP Address
Options
Padding
Protocol: specifies upper-layer protocol that is to receive IP data at the destination. Examples include TCP (protocol = 6), UDP (protocol = 17), and ICMP (protocol = 1).
Header checksum: verifies the integrity of the IP header.
Source IP address and destination IP address: contain the addresses of the source and destination hosts.
IP Packet Header
048 161924 31
Version
IHL
Type of Service
Total Length
Identification
Flags
Fragment Offset
Time to Live
Protocol
Header Checksum
Source IP Address
Destination IP Address
Options
Padding
Options: Variable length field, allows packet to request special features such as security level, route to be taken by the packet, and timestamp at each router. Detailed descriptions of these options can be found in [RFC 791].
Padding: This field is used to make the header a multiple of 32-bit words.
Example of IP Header
Header Checksum
IP header uses check bits to detect errors in the header
A checksum is calculated for header contents
Checksum recalculated at every router, so algorithm selected for ease of implementation in software
IP Header Processing
1. Compute header checksum for correctness and check that fields in header (e.g. version and total length) contain valid values
2. Consult routing table to determine next hop
3. Change fields that require updating (TTL, header checksum)
Non-Hierarchical Addresses and Routing
0000 0111 1010 1101
0011 0110 1001 1100
1
R1 25
0001 0100 1011 1110
0011 0101 1000 1111
4 3
R2
0001 0100 1011 …
4 4 4 …
0000 0111 1010 …
1 1
1
…
No relationship between addresses & routing proximity
Routing tables require 16 entries each
Hierarchical Addresses and Routing
0000 0001 0010 0011
1000 1001 1010 1011
1
4 3
0100 0101 0110 0111
1100 1101 1110 1111
R2 25
R1
00 01 10 11
1 3 2 3
00 01 10 11
3 4 3 5
Prefix indicates network where host is attached
Routing tables require 4 entries each
IP Addressing
RFC 1166
Each host on Internet has unique 32 bit IP address
Each address has two parts: netid and hostid
netid unique & administered by
American Registry for Internet Numbers (ARIN) Reseaux IP Europeens (RIPE)
Asia Pacific Network Information Centre (APNIC)
Facilitates routing
A separate address is required for each physical
connection of a host to a network; “multi-homed” hosts
Dotted-Decimal Notation:
int1.int2.int3.int4 where intj = integer value of jth octet IP address of 10000000 10000111 01000100 00000101 is 128.135.68.5 in dotted-decimal notation
Classful Addresses
Class A
7 bits 24 bits
0
netid
hostid
• 126 networks with up to 16 million hosts
1.0.0.0 to 127.255.255.255
Class B
14 bits
16 bits
1
0
netid
hostid
• 16,382 networks with up to 64,000 hosts
128.0.0.0 to 191.255.255.255
Class C
22 bits
8 bits
1
1
0
netid
hostid
• 2 million networks with up to 254 hosts
192.0.0.0 to 223.255.255.255
Class D
28 bits
1
1
1
0
multicast address
224.0.0.0 to 239.255.255.255
Up to 250 million multicast groups at the same time
Permanent group addresses
All systems in LAN; All routers in LAN;
All OSPF routers on LAN; All designated OSPF routers on a LAN, etc.
Temporary groups addresses created as needed
Reserved Host IDs (all 0s & 1s)
Internet address used to refer to network has hostid set to all 0s
0
0
0
0
0
0
0
0
0
host
Broadcast address has hostid set to all 1s
this host (used when booting up)
a host in this network
broadcast on local network
broadcast on distant network
1
1
1
1
1
1
netid
1
1
1
1
1
1
1
Private IP Addresses
Specific ranges of IP addresses set aside for use in private networks (RFC 1918)
Use restricted to private internets; routers in public Internet discard packets with these addresses
Range 1: 10.0.0.0 to 10.255.255.255
Range 2: 172.16.0.0 to 172.31.255.255
Range 3: 192.168.0.0 to 192.168.255.255
Network Address Translation (NAT) used to convert between private & global IP addresses
Example of IP Addressing
128.135.40.1
128.140.5.40
Interface Address is
128.135.10.2
Interface Address is
128.140.5.35
H
H
Network 128.135.0.0
Network 128.140.0.0
R
H
128.135.10.20 128.135.10.21
128.140.5.36
R = router H = host
H
Address with host ID=all 0s refers to the network Address with host ID=all 1s refers to a broadcast packet
H
Subnet Addressing
Subnet addressing introduces another hierarchical level
Transparent to remote networks
Simplifies management of multiplicity of LANs
Masking used to find subnet number
Original address
Subnetted address
1
0
Net ID
Host ID
1
0
Net ID
Subnet ID
Host ID
Subnetting Example
Organization has Class B address (16 host ID bits) with network ID: 150.100.0.0
Create subnets with up to 100 hosts each 7 bits sufficient for each subnet
16-7=9 bits for subnet ID
Apply subnet mask to IP addresses to find corresponding subnet
Example: Find subnet for 150.100.12.176
IP add = 10010110 01100100 00001100 10110000 Mask = 11111111 11111111 11111111 10000000
AND = 10010110 01100100 00001100 10000000 Subnet = 150.100.12.128
Subnet address used by routers within organization
Subnet Example
H1
150.100.12.154 150.100.12.129
150.100.0.1 R1
150.100.12.176
To the rest of the Internet
150.100.12.4H3 150.100.12.24
150.100.12.0
H2
150.100.12.128
H4
150.100.12.55 150.100.12.1
R2
150.100.15.54 150.100.15.11
H5
150.100.15.0
Routing with Subnetworks
IP layer in hosts and routers maintain a routing table
Originating host: To send an IP packet, consult
routing table
If destination host is in same network, send packet directly using appropriate network interface
Otherwise, send packet indirectly; typically, routing table indicates a default router
Router: Examine IP destination address in arriving packet
If dest IP address not own, router consults routing table to determine next-hop and associated network interface & forwards packet
Routing Table
Each row in routing table contains:
Destination IP address
IP address of next- hop router
Physical address
Statistics information
Flags
H=1 (0) indicates route is
to a host (network)
G=1 (0) indicates route is to a router (directly connected destination)
Routing table search order & action
Complete destination address; send as per next- hop & G flag
Destination network ID; send as per next-hop & G flag
Default router entry; send as per next-hop
Declare packet undeliverable; send ICMP “host unreachable error” packet to originating host
Example: Host H5 sends packet to host H2
150.100.12.154 150.100.12.128
150.100.12.129
150.100.12.4H3 150.100.12.24
150.100.12.176
150.100.0.1
To the rest of the Internet
Destination 127.0.0.1 default 150.100.15.0
Next-Hop 127.0.0.1 150.100.15.54 150.100.15.11
Flags Net I/F H lo0
G emd0
emd0
R1
H1
150.100.12.0
150.100.12.55 150.100.12.1
Routing Table at H5
150.100.15.54 150.100.15.0
150.100.12.176
150.100.15.11
R2
H2
H4
H5
Example: Host H5 sends packet to host H2
150.100.0.1
To the rest of the Internet
Routing Table at R2
150.100.12.55 150.100.12.1
150.100.12.154 150.100.12.128
150.100.12.129
150.100.12.4H3 150.100.12.24
150.100.12.176
Destination 127.0.0.1 default 150.100.15.0 150.100.12.0
Next-Hop 127.0.0.1 150.100.12.4 150.100.15.54 150.100.12.1
Flags Net I/F H lo0
G emd0
emd1 emd0
R1
150.100.12.0
150.100.12.176
H1
R2
H2
H4
150.100.15.54 150.100.15.11
H5
150.100.15.0
Example: Host H5 sends packet to host H2
150.100.12.154 150.100.12.176 150.100.12.128
150.100.12.129 150.100.12.176
150.100.0.1
To the rest of the Internet
Routing Table at R1
150.100.12.4H3 150.100.12.24
Destination 127.0.0.1 150.100.12.176 150.100.12.0 150.100.15.0
Next-Hop 127.0.0.1 150.100.12.176 150.100.12.4 150.100.12.1
Flags H
G
Net I/F lo0 emd0 emd1 emd1
R1
H1
R2
H2
150.100.12.0
150.100.12.55 150.100.12.1
150.100.15.54 150.100.15.11
H4
H5
150.100.15.0
IP Address Problems
In the 1990, two problems became apparent IP addresses were being exhausted
IP routing tables were growing very large
IP Address Exhaustion
Class A, B, and C address structure inefficient
Class B too large for most organizations, but future proof Class C too small
Rate of class B allocation implied exhaustion by 1994
IP routing table size
Growth in number of networks in Internet reflected in # of table entries
From 1991 to 1995, routing tables doubled in size every 10 months Stress on router processing power and memory allocation
Short-term solution:
Classless Interdomain Routing (CIDR), RFC 1518
New allocation policy (RFC 2050)
Private IP Addresses set aside for intranets
Long-term solution: IPv6 with much bigger address space
New Address Allocation Policy
Class A & B assigned only for clearly demonstrated need
Consecutive blocks of class C assigned (up to 64 blocks)
All IP addresses in the range have a common prefix, and every address with that prefix is within the range
Arbitrary prefix length for network ID improves efficiency
Address Requirement
Address Allocation
< 256
1 Class C
256<,<512
2 Class C
512<,<1024
4 Class C
1024<,<2048
8 Class C
2048<,<4096
16 Class C
4096<,<8192
32 Class C
8192<,<16384
64 Class C
Supernetting
Summarize a contiguous group of class C addresses using variable-length mask
Example: 150.158.16.0/20
IP Address (150.158.16.0) & mask length (20)
IP add = 10010110 10011110 00010000 00000000 Mask = 11111111 11111111 11110000 00000000
Contains 16 Class C blocks:
From 10010110 10011110 00010000 00000000 i.e. 150.158.16.0
Up to 10010110 10011110 00011111 00000000 i.e. 150.158.31.0
Classless Inter-Domain Routing
CIDR deals with Routing Table Explosion Problem
Networks represented by prefix and mask
Pre-CIDR: Network with range of 16 contiguous class C blocks requires 16 entries
Post-CIDR: Network with range of 16 contiguous class C blocks requires 1 entry
Solution: Route according to prefix of address, not class
Routing table entry has
Example: 192.32.136.0/21
11000000 00100000 10001000 00000001 min address 11111111 11111111 11111— ——– mask
11000000 00100000 10001— ——– IP prefix
11000000 00100000 10001111 11111110 max address 11111111 11111111 11111— ——– mask
11000000 00100000 10001— ——– same IP prefix
Longest Prefix Match
CIDR impacts routing & forwarding
Routing tables and routing protocols must carry IP
address and mask
Multiple entries may match a given IP destination address
Example: Routing table may contain
205.100.0.0/22 which corresponds to a given supernet
205.100.0.0/20 which results from aggregation of a larger number of destinations into a supernet
Packet must be routed using the more specific route, that is, the longest prefix match
Several fast longest-prefix matching algorithms are available
Address Resolution Protocol
Although IP address identifies a host, the packet is physically delivered by an underlying network (e.g., Ethernet) which uses its own physical address (MAC address in Ethernet). How to map an IP address to a physical address?
H1 wants to learn physical address of H3 -> broadcasts an ARP request
H1
H2
H3
150.100.76.22
ARP request (what is the MAC address of 150.100.76.22?)
150.100.76.23
150.100.76.20 150.100.76.21
Every host receives the request, but only H3 reply with its physical address
ARP response (my MAC address is 08:00:5a:3b:94)
H4
H1
H2
H3
H4
Example of ARP
Internet Control Message Protocol (ICMP)
RFC 792; Encapsulated in IP packet (protocol type = 1)
Handles error and control messages
If router cannot deliver or forward a packet, it sends an ICMP “host unreachable” message to the source
If router receives packet that should have been sent to another router, it sends an ICMP “redirect” message to the sender; Sender modifies its routing table
ICMP “router discovery” messages allow host to learn about routers in its network and to initialize and update its routing tables
ICMP echo request and reply facilitate diagnostic and used in “ping”
UDP
Best effort datagram service
Multiplexing enables sharing of IP datagram service
Simple transmitter & receiver
Connectionless: no handshaking & no connection state Low header overhead
No flow control, no error control, no congestion control UDP datagrams can be lost or out-of-order
Applications
multimedia (e.g. RTP)
network services (e.g. DNS, RIP, SNMP)
UDP Datagram
0 16 31
Source and destination port numbers
Client ports are short-lived Server ports are well-known Max number is 65,535
UDP length
Total number of bytes in
datagram (including header)
8 bytes ≤ length ≤ 65,535
UDP Checksum
Optionally detects errors in UDP datagram
Source Port
Destination Port
UDP Length
UDP Checksum
Data
0-255
Well-known ports
256-1023
Less well-known ports
1024-65536
Client ports
UDP Multiplexing
All UDP datagrams arriving to IP address B and destination port number n are delivered to the same process
Source port number is not used in multiplexing
1
2 … n 1 2 … n 1 2 … n
UDP
UDP
UDP
A
BC
IP
IP
IP
TCP
Reliable byte-stream service
More complex transmitter & receiver
Connection-oriented: full-duplex unicast connection between client & server processes
Connection setup, connection state, connection release
Higher header overhead
Error control, flow control, and congestion control
Higher delay than UDP
Most applications use TCP
HTTP, SMTP, FTP, TELNET, POP3, …
Reliable Byte-Stream Service
Stream Data Transfer
transfers a contiguous stream of bytes across the network,
with no indication of boundaries
groups bytes into segments
transmits segments as convenient (Push function defined)
Reliability
error control mechanism to deal with IP transfer impairments
Application Transport
Write 45 bytes Write 15 bytes Write 20 bytes
Read 40 bytes Read 40 bytes
buffer
Error Detection & Retransmission
buffer
segments ACKS, sequence #
Flow Control
Buffer limitations & speed mismatch can result in loss of data that arrives at destination
Receiver controls rate at which sender transmits to prevent buffer overflow
Application Transport
segments
advertised window size < B
buffer used buffer available = B
buffer
Congestion Control
Available bandwidth to destination varies with activity of other users
Transmitter dynamically adjusts transmission rate according to network congestion as indicated by RTT (round trip time) & ACKs
Elastic utilization of network bandwidth Application
Transport
RTT Estimation
buffer
segments ACKS
buffer
TCP Multiplexing
A TCP connection is specified by a 4-tuple
(source IP address, source port, destination IP address,
destination port)
TCP allows multiplexing of multiple connections between end systems to support multiple applications simultaneously
Arriving segment directed according to connection 4-tuple
1
2 ... m
1 2 ... n 1 2 ... k
TCP
TCP
TCP
A
(A, 6234, B, 80) (A, 5234, B, 80)
BC (C, 5234, B, 80)
IP
IP
IP
TCP Segment Format
0 4 10 16 24 31
Source port
Destination port
Sequence number
Acknowledgment number
Header length
Reserved
U R G
A C K
P S H
R S T
S Y N
F I N
Window size
Checksum
Urgent pointer
Options
Padding
Data
• Each TCP segment has header of 20 or more bytes + 0 or more bytes of data
TCP Header
Port Numbers
A socket identifies a connection endpoint
IP address + port
A connection specified
by a socket pair
Well-known ports FTP 20
Telnet 23 DNS 53 HTTP 80
Sequence Number
Byte count
First byte in segment 32 bits long
0 SN 232-1
Initial sequence number selected during connection setup
TCP Header
Acknowledgement Number
SN of next byte expected by receiver
Acknowledges that all prior bytes in stream have been received correctly
Valid if ACK flag is set
Header length
4 bits
Length of header in multiples of 32-bit words
Minimum header length is 20 bytes
Maximum header length is 60 bytes
TCP Header
Reserved
6 bits
Control
6 bits
URG: urgent pointer flag
Urgent message end = SN + urgent pointer
ACK: ACK packet flag
PSH: override TCP buffering
RST: reset connection
Upon receipt of RST, connection is
terminated and application layer notified SYN: establish connection
FIN: close connection
TCP Header
Window Size
16 bits to advertise window size
Used for flow control
Sender will accept bytes with SN from ACK to ACK + window
Maximum window size is 65535 bytes
TCP Checksum
Internet checksum method
TCP Header
Options Options
Variable length
NOP (No Operation) option is used to pad TCP header to multiple of 32 bits
Time stamp option is used for round trip measurements
Maximum Segment Size (MSS) option specifies largest segment a receiver wants to receive
Window Scale option increases TCP window from 16 to 32 bits
Initial Sequence Number
Select initial sequence numbers (ISN) to protect against segments from prior connections (that may circulate in the network and arrive at a much later time)
Select ISN to avoid overlap with sequence numbers of prior connections
Use local clock to select ISN sequence number
Time for clock to go through a full cycle should be greater than the maximum lifetime of a segment (MSL); Typically MSL=120 seconds
High bandwidth connections pose a problem
32bit SN wraps around after 232 = 4.29x109 bytes = 34.3x109 bits have been sent
At 1 Gbps, sequence number wraparound in 34.3 seconds.
TCP Connection Establishment
• “Three-way Handshake”
• ISN’s protect against segments
Host A from prior connections Host B
If host always uses the same ISN
Host A Host B
Delayed segment with Seq_no = n+2
will be accepted
Maximum Segment Size
Maximum Segment Size
largest block of data that TCP sends to other end
Each end can announce its MSS during connection establishment
Default is 576 bytes including 20 bytes for IP header and 20 bytes for TCP header
Ethernet implies MSS of 1460 bytes
IEEE 802.3 implies 1452
Near End: Connection Request
Far End: Ack and Request
Near End: Ack
TCP Window Flow Control
Host A
Host B
t0
1024 bytes to transmit
1024 bytes t1 to transmit
t2 1024 bytes
1024 bytes to transmit
t4
128 bytes to transmit
t3
can only send 512 bytes
to transmit
Delay-BW Product & Advertised Window Size
Suppose RTT=100 ms, R=2.4 Gbps # bits in pipe = 3 Mbytes
If single TCP process occupies pipe, then required advertised window size is
RTT x Bit rate = 3 Mbytes
Normal maximum window size is 65535 bytes
Solution: Window Scale Option
Window size up to 65535 x 214 = 1 Gbyte allowed Requested in SYN segment
TCP Connection Closing
Host A
“Graceful Close”
Host B
Deliver 150 bytes
TCP Congestion Control
Advertised window size is used to ensure that receiver’s buffer will not overflow
However, buffers at intermediate routers between source and destination may overflow
Packet flows from many sources
Router
R bps
Congestion occurs when total arrival rate from all packet flows exceeds R over a sustained period of time
Buffers at multiplexer will fill and packets will be lost