Transcript Broadcast Address

DCN286 INTRODUCTION TO DATA
COMMUNICATION TECHNOLOGY
TCP/IP and IP addressing
Internet and TCP/IP
• Internet is a global network of interconnected computers,
enabling users to share information along multiple channels.
• Majority of widely accessible information on the internet
consists of inter-linked hypertext documents and other
resources of the World Wide Web (WWW).
• The Internet Protocol Suite (commonly known as TCP/IP) is
the set of communications protocols used for the Internet and
other similar networks. It is named from two of the most
important protocols in it: the Transmission Control Protocol
(TCP) and the Internet Protocol (IP), which were the first two
networking protocols defined in this standard.
• TCP/IP is an open standard under direction of Internet
Architecture Board (IAB) and Internet Engineering Task Force
(IETF).
TCP/IP model
Application
Application
Presentation
Session
Transport
Transport
Internet
Network
Network Interface
Data Link
Physical
TCP/IP protocol suite
TCP/IP Network Access Layer
• Defines how computers and network devices should
access a physical medium to send bits to another
computer.
• Network Access layer has two ISO OSI layers, layer 1
and 2 (Physical and Data Link)
• IETF does not define any standard in this layer because
other organizations define many standards. For instance,
Ethernet is developed by IEEE.
• PDU (Protocol Data Unit) is bit in layer 1 (Physical) and
frame in layer 2 (Data Link).
TCP/IP Internet Layer
•
1.
2.
3.
4.
Defines how to deliver data from one host to
another on various physical networks:
Logical addressing
Routing and routing protocol
Main protocols: IP, ARP, RARP (Reverse Address
Resolution Protocol), ICMP (Internet Control
Message Protocol) and router protocols such as
RIP (Routing Information Protocol) and OSPF
(Open Shortest Path First).
PDU (Protocol Data Unit) is packet in layer 3
ICMP (Internet Control Message
Protocol)
ICMP can report error for diagnostic or
routing control purpose. Two key
commands are used to test connectivity:
• PING (ICMP echo request and reply
command)
• Tracert is Microsoft version command
while Cisco uses traceroute to display
routing path while testing connectivity.
TCP/IP Host-to-Host (Transport) Layer
• Two main protocols: TCP (Transmission Control
Protocol) and UDP (User Datagram Protocol)
• TCP provides reliable data delivery service with end-toend error detection and correction.
• UDP provides low-overhead, connection less datagram
delivery services.
• Application programmers can choose which protocol to
use.
• PDU (Protocol Data Unit) is segment in layer 4
(Transport)
TCP and UDP
The two important transport layer protocols:
They help to transfer data from one application
process on one computer to the correct
application process on another computer.
• TCP (Transmission Control Protocol)
Connection-oriented protocol, more reliable and
slower.
• UDP (User Datagram Protocol)
Connection-less, faster, simpler and not reliable
(no error checking).
TCP and UDP comparison
Features
Flow control and windowing
TCP UDP
Yes
No
Connection-oriented
Yes
No
Error recovery
Yes
No
Segmentation and reassembly of data
Yes
No
In-order delivery of data
Yes
No
Identifying application using port
numbers
Yes
Yes
TCP/IP Application Layer
Offer interface to deal with various
applications:
• File transfer: FTP, FTP, NFS
• Email: SMTP, POP3, IMAP
• Remote login: Telnet
• Network Management: SNMP, DNS
• Web browsing: HTTP, HTTPS
Some protocols (1)
•
•
•
•
•
•
•
•
•
•
•
FTP (File Transfer Protocol)
TFTP (Trivial File Transfer Protocol)
NFS (Network File System)
SMTP (Simple Mail Transfer Protocol)
LPD (Line Printer Daemon)
SNMP (Simple Network Management
Protocol)
DNS (Domain Name Service)
DHCP (Dynamic Host Control Protocol)
TCP (Transmission Control Protocol)
UDP (User Datagram Protocol)
IP (Internet Protocol)
Some protocols (2)
• ARP (Address Resolution Protocol)
• RARP (Reverse Address Resolution
Protocol)
• POP3 (Post Office Protocol V3)
• IMAP (Internet Message Access Protocol)
• RIP (Routing Information Protocol)
• OSPF (Open Shortest Path First)
IPV4 address
1. IP address must be unique in the network (Public IP
address must be unique in internet and private IP
address must be unique in the internal network)
2. The format is xxxxxxxx.xxxxxxxx.xxxxxxxx.xxxxxxxx
(totally 32-bit).There are public IP address used in
Internet and private IP address used in
company/organization internal networks.
3. As 8 bits binary can present 0-255 in decimal, IP
address could be (and usually is) presented as
10.1.11.12 (You must not see such IP address
260.270.280.290)
4. Network mask is to divide the IP address into two
logical sections (network ID and host ID)
Network Mask
• IP address has both Network ID (network address)
and Host ID (host address). If the value is 1 in the
mask, it means the corresponding bit in the IP
address is the network ID. If the value is 0, it means
the corresponding bit is for the host ID. For instance,
network musk is 255.0.0.0 which equals to binary
value 11111111.00000000.00000000.00000000.
The “1” section presents the network ID. So,
10.10.10.10 would have the first octet 10 as network
ID. If the network mask is 255.255.255.0, the
network ID of 192.168.10.10 would be 192.168.10.0
because 255.255.255 is
11111111.11111111.11111111 in the mask.
Valid network mask
Binary value
00000000
10000000
11000000
11100000
11110000
11111000
11111100
11111110
11111111
Decimal value
0
128
192
224
240
248
252
254
255
IP address class
Traditionally, the IP address was classified in classes:
Class A: network 1 – 126 with subnet mask 255.0.0.0
(Initial bit starts as 0)
Class B: network 128 – 191 with subnet mask
255.255.0.0 (Initial bit starts as 10)
Class C: network 192 – 223 with subnet mask
255.255.255.0 (Initial bit starts as 110)
Class D: network 224 – 239 (Initial bit starts as 1110)
for multicast IP address
Class E: network 240 – 255 (Initial bit starts as 1111)
experimental use
The special network 0 and127 are not included in
those official classical IP address. 127.0.0.1 is the
loopback address which is used to test TCP/IP
stacks.
Some basic rules
• In same network, no routing or router is
required.
• In same network, IP address must be unique
to identify the computer or network host. In
the internet, public IP address must be
globally unique (we will introduce private IP
address soon).
• Router or routing is required between different
networks.
Power calculation of 2
The powers of 2
Decimal value
Binary Value
2^0
1
00000001
2^1
2
00000010
2^2
4
00000100
2^3
8
00001000
2^4
16
00010000
2^5
32
00100000
2^6
64
01000000
2^7
128
10000000
2^8
256
100000000
2^9
512
1000000000
2^10
1024
10000000000
2^11
2048
100000000000
2^12
4096
1000000000000
2^13
8192
10000000000000
Network address and Broadcast
address
• Network address is also called network ID. It is
the part defined by subnet mask. For instance,
network musk is 255.0.0.0 which equals to binary
value 11111111.00000000.00000000.00000000.
The “1” section presents the network ID. So,
10.10.10.10 would have the first octet 10 as
network ID. Network ID is to have all 0 in the host
ID section.
• If the broadcast address is the destination
address of a packet, it would be sent to all hosts
in the same network. Broadcast address is to
have all 1 in the host ID section. For instance,
broadcast address of 10.10.10.10/8 is
10.255.255.255
Unicasts, Multicasts, and Broadcasts
Unicasts:
– Packet travels from one host to another specific host.
Multicasts:
– Packet travels from one host to a select number of
other hosts.
– Supports voice and audio broadcasts, news feeds,
distribution of
software, re-imaging clients off peak times.
Broadcasts:
– Packet travels from one host to all hosts on the local
network.
Broadcast Address
All of the bits in the host portion of the
broadcast IP address ends with binary 1s.
– For network 176.10.0.0
•10110000.0001010.00000000.00000000
– The broadcast address is
176.10.255.255
So in binary----10110000.0001010.11111111.11111111
Valid host number
• If the host id bits are all 0 in IP address, it presents the network address; If
the host id bits are all 1 in IP address, it presents the broadcast address of
the network. Valid host cannot have those 2 special (reserved) IP address.
• Valid host number is: 2^(number-of-host-bits) – 2
• Computer operating systems usually would not treat 192.168.4.0 as a valid
host IP address. For the same reason, it cannot use 192.168.4.255 as the
computer IP address.
Class Size of host
id (bytes)
A
3
B
C
2
1
Size of host Number of hosts
id (bits)
per network
24
2^24 – 2 =
16,777,214
16
2^16 – 2 = 65,534
8
2^8 – 2 =254
Network number calculation
The tricky part is that there are fixed bits in network
ID of all classes. Class A has 0 in first bit all the
time (0-126). Class B always has 10 in the first two
bits. Class C has 110 as fixed bits.
Class Size of
network id
(bytes)
Size of
network
id (bits)
Number of
fixed bits at
beginning
Number of
network bits
that vary
Number of
networks
A
1
8
1
7
2^7 – 2 = 126
B
2
16
2
14
2^14 – 2 =
16,192
C
3
24
3
21
2^21 – 2 =
2,097,152
Subnet
Subnet is to logically divide your network into many
sub networks.
• In the same subnet, traffic is “local” and not gateway
(router) is required. Network hosts would use ARP
table for the MAC address of the destination machine
and send the packet to it accordingly. If it is not local,
packets would be forwarded to default gateway for
future routing (redirecting).
• In addition, the broadcast would be limited to the small
subnets leading to less “noise” in the network traffic.
• You can also use subnet to logically specify hosts for
different departments. (for security control, connection,
asset management, etc, etc)
Define subnet number
• Subnet bits (“1”) can tell how many subnet could be produced by the
mask. The formula is 2^(number of subnet “1” bits) - 2
Original classfull network mask:
• Class A network has the form N.H.H.H, the default subnet mask is 8 bits
long.
• Class B network has the form N.N.H.H, the default subnet mask is 16
bits long.
• Class C network has the form N.N.N.H, the default subnet mask is 24
bits long.
The additional subnet bits can generate subnet. For instance, You have an
IP of 156.233.0.0 with a subnet mask of 7 bits. How many hosts and
subnets are possible?
There is additional 7 bits to the default subnet mask. The total number of
bits in subnet are 16+7 = 23. This leaves us with 32-23 =9 bits for
assigning to hosts. 7 bits of subnet mask corresponds to (2^7-2)=128-2
= 126 subnets. 9 bits belonging to host addresses correspond to (2^92)=512-2 = 510 hosts.
Define host number in each subnet
• Subnet bits (“0”) can tell how many host could be contained in each
subnet. The formula is 2^(number of host “0” bits) - 2
Original classfull network mask:
• Class A network has the form N.H.H.H, the default host mask is 24 bits
long.
• Class B network has the form N.N.H.H, the default host mask is 16 bits
long.
• Class C network has the form N.N.N.H, the default host mask is 8 bits
long.
The additional subnet bits can generate subnet. For instance, You have an
IP of 156.233.0.0 with a subnet mask of 7 bits. How many hosts and
subnets are possible?
There is additional 7 bits to the default subnet mask. The total number of
bits in subnet are 16+7 = 23. This leaves us with 32-23 =9 bits for
assigning to hosts. 7 bits of subnet mask corresponds to (2^7-2)=128-2
= 126 subnets. 9 bits belonging to host addresses correspond to (2^92)=512-2 = 510 hosts.
Routing
• If all traffics are in the same LAN (subnet),
no routing is required. The computers talk
to each other over network cable directly.
• In different LAN (subnet), the packet sent
from sender must be “redirected” (routed)
to another network in which the recipient is
located.
Cisco 2610 interfaces (optional)
Public IP address
IANA (Internet Assigned Numbers Authority
www.iana.org) has the ultimate authority. It
authorizes member organizations to assign
registered unique IP address to users. For
instance, ARIN (American Registry for Internet
Numbers) is in charge of assigning IP address in
US.
In general, each computer needs a public IP
address to identify itself and access Internet.
Private IP address
Certain blocks are reserves as private IP address used in company /
organization internal networks. More computers could have “unique” IP
address in IPV4.
Class Range of IP
address
Range of
Network number
Default network
mask
Total number
of networks
1
A
10.0.0.0 –
10.255.255.255
10.0.0.0 –
10.0.0.0
/8 (255.0.0.0)
B
172.16.0.0 –
172.31.255.255
172.16.0.0 –
172.31.255.255
/12 (255.240.0.0) 16
C
192.168.0.0 –
192.168.0.0 –
/16 (255.255.0.0) 256
192.168.255.255 192.168.255.255
Thus, it is possible for many companies / organizations using same
network address internally. (For instance, 10.10.10.10 could be used for
computer Tom in company Apple and by computer Jerry in company
Boss) But, again, the potential duplicated addresses are used internally.
NAT (Network Address Translation)
• Allow 65,535 user connections with the same public
(registered) IP address.
• Enable multiple users (computers) sharing single
internet connection.
• The computers may have internal private IP
address. But, they can have public identifier via the
public IP address.
• The solution is to use NAT map to associate user
connection from internal private IP address and
public IP address. The NAT map is beyond this
course. But, port number is used to have the manyto-many mapping.
NAT Illustration (optional)
IPV6
•
•
•
•
128 bits address format
6 octets
Usually presented in hex
May support 10^30 IP addresses (IPV4 only
supports 4,294,467,295 IP addresses)
• Still not in final implementation
• NOT interoperable with IPV4
IP address assignment
IP address could be assigned by:
• Static
User manually assign the IP address, mask,
default gateway and DNS.
• Dynamic (automatically) by DHCP (Dynamic
Host Control Protocol) server
No need to do any configuration. Enable DHCP
client and the computer (client) would get IP
address and other info from DHCP server.
(Network Administrator does need to do
configuration on server side.)
Compare static and DHCP
Good for large network
Client side configuration
Server side configuration
Move whole network address to
another number blocks (change
172.20.1.0 to 192.168.5.0)
Can reuse the IP address by other
computers
Security control
DHCP
Yes
No
Yes
Static
No
Yes
No
Yes
No
Yes
No
Average High
ARP (Address Resolution Protocol)
•
ARP table is to contain IP address and
associated MAC address information.
• To add or update ARP table:
1. Senders sends ARP Request (ARP
broadcast) out.
2. Receiver returns ARP Reply (unicast)
back.
• RARP is to offer IP address against MAC
address (reverse direction)
ARP command
• ARP command can be used to display,
add and clear entries in ARP table:
Arp /? can display all options of the
command:
arp –a
arp –s
arp -d
Question
Any question?
If you do not have question, please start
subnet calculation.
1. What subnet mask would be if I want to have
10 subnets in 10.0.0.0/8?
2. What subnet mask would be if I want to have
20 hosts in each subnet in 192.169.12.0/24?
Do not forget the lab assignment.