SUBNET

1
SUBNET ROUTING

• Sritrusta Sukaridhoto

2
IP Addressing
3
IP addressing

• Introduction to IP addressing
• Classes of IP addressing
• Why Subnet Masks are necessary?
• How to create subnet masks

4
Why are IP addresses written as bits?

• In order for data to pass along the media, it
  must first be changed to electrical impulses.
• When a computer receives these electrical
  impulses, it recognizes two things the presence
  of voltage on the wire or the absence of voltage
  on the wire.

5
(No Transcript)
6
What format do IP addresses use?
7
How are IP addresses expressed in dotted notation?
8
Why are IP addresses necessary?

• IP addressing makes it possible for data passing
  over the network media of the Internet to find
  its destination.
• Because each IP address is a 32-bit value, that
  means that there are four billion different IP
  address possibilities.
• IP addresses are hierarchical addresses like
  phone numbers and zip codes.

9
How do IP addresses make it possible for data
sent via the Internet to find its destination?

• It is because each network connected to the
  Internet has a unique network number.
• To ensure that each network number on the
  Internet will always be unique and unlike that of
  any other number, an organization called the
  International Network Information Center, or
  InterNIC

10
How do IP addresses incorporate network
addresses?

• Every IP address has two parts. These are known
  as the network number and the host number.

11
What are the different classes of networks?

• There are three classes of IP addresses that a
  company can receive from the InterNIC. The
  InterNIC reserves class "A" IP addresses for
  governments throughout the world, class "B" IP
  addresses for medium size companies, and class
  "C"

12
IP addressing

• Classes of Networks

1 Byte
1 Byte
1 Byte
1 Byte
Class A
Class B
Class C
Network number assigned by NICHost number
assigned by Systems Administrator
13
IP addressing

• IP Address Bit Patterns

1
2 - 8
9 32 (24 bits)
Bit
Class A
Class A address range1.0.0.0 126.0.0.0
(127.0.0.0 is for loopback)Private Class A
address 10.0.0.0Number of hosts 224 -2
16,777,214
14
IP addressing

• IP Address Bit Patterns

1
3 - 16
17 32 (16 bits)
2
Bit
Class B
Class B address range128.0.0.0
191.255.0.0Private Class B 172.16.0.0
172.31.0.0 Number of hosts 216 - 2 65,534
15
IP addressing

• IP Address Bit Patterns

1
4 - 24
25 32 (8 bits)
2
3
Bit
Class C
Class C address range192.0.0.0
233.255.255.0Private Class C 192.168.0.0
Number of hosts 28 - 2 254
16
(No Transcript)
17
How many classes of Networks are there?

• you have learned about three classes of networks
  that can be assigned by the InterNIC.
• In fact, there are five classes of networks.
  However, only three of these are used
  commercially.

18
What IP addresses are reserved for multicast
purposes and experimental purposes?

• The highest number listed was 223. You may have
  wondered why the highest value was only 223 and
  not 255, since there are 255 possible values for
  an octet.
• in IP addresses the values 224 through 255 are
  not used in the first octet for networking
  purposes.

19
What IP addresses are reserved for the Networks?

• By convention, in IP addressing schemes, any IP
  address that ends in all binary zeroes is
  reserved for the networkaddress.
• Thus, in a class "A" network, 113.0.0.0 would be
  the IP address of that network. Routers use a
  network's IP address when forwarding data on the
  Internet.

20
(No Transcript)
21
What IP addresses are reserved for broadcasts?

• for the network that is 176.10.0.0, the broadcast
  address that would be sent out to all devices on
  that network would be 176.10.255.255.

22
Who assigns subnet addresses?

• As with the host number portion of class "A,"
  class "B," and class "C addresses, subnet
  addresses are assigned locally.
• Usually this is done by the network administrator.

23
(No Transcript)
24
IP addressing Subnet mask

• Subnet Masking - Why?

IP address
Netmask
Network Address


25
IP addressing - Subnetting

• Subnetting
• Is the act of borrowing bits from the host
  portion to create smaller networks (called
  subnetworks)
• Minimum bits that can be borrowed is 2 - why?
• Subnetting is used to reduce the number of
  broadcast domains
• Communication between these subnetworks is
  achieved through a router

26
(No Transcript)
27
How are subnet addresses concealed from outside
networks?

• Subnets are hidden from outside networks by using
  a mask.
• These are referred to as subnet masks.
• The function of a subnet mask is to tell devices
  which part of an address is the network number
  including the subnet, and which part is the host.
  

28
What format do subnet masks use?

• Subnet masks use the same format as IP
  addressing.
• In other words, they are thirty two bits long and
  divided into four octets.
• Subnet masks have all 1s in the network and
  subnetwork portion, and all 0s in the host
  portion.

29
(No Transcript)
30
(No Transcript)
31
How many bits can be borrowed from the host
number in class "B" and class "C" networks to
create subnets?

• Because there are only two octets in the host
  field of a class "B network, up to fourteen bits
  can be borrowed to create subnetworks.
• A class "C" network has only one octet in the
  host field. Therefore, only up to six bits can be
  borrowed in class "C networks to create
  subnetworks.

32
What happens to the subnet mask address if only
some of the bits in an octet are borrowed?

• Imagine that you have a class "B network. This
  time however, instead of borrowing all eight bits
  of the third octet, only seven bits are borrowed
  to create subnetworks.
• Using binary representation, in this example, the
  subnet mask would be 11111111.11111111.11111110.00
  000000.
• Therefore, 255.255.255.0 can no longer be used as
  the subnet mask.

33
If only seven bits are borrowed in a class "B"
network, what would the subnet mask be in dotted
decimal notation?

• HINT To convert any eight bit binary number into
  a decimal number, total the powers of 2 that
  occur in the number.

34
What determines how many subnetworks can be
created by borrowing bits from the host field?

• Can you figure out all of the possible
  combinations of 0s and 1s if four bits are
  borrowed from the host field to create
  subnetworks?
• 16 from 0000 to 1111. However, you know that 1111
  is reserved for broadcast and 0000 means this
  network.

35
How many subnetworks can be created by borrowing
five bits from the host field?

• Answer Thirty-two subnetworks or 25 32
  subnetworks can be created by borrowing five
  bits from the host field.

36
Which numbers in a subnetwork are reserved for
broadcasts?

• In previous section, we used an example of a
  class "C" network in which three bits are
  borrowed from the host field. You learned that
  when three bits are borrowed from the host octet,
  up to eight subnetworks can be created each
  having up to thirty-two hosts.
• You also learned that IP addresses ending in all
  binary 1s are reserved for broadcasts. The same
  is true for subnetworks.

37
For IP address 197.15.22.160 on the same class
"C network. How would this be expressed in a
binary numbering scheme?

• Answer If 197.15.22.160 is converted to binary
  format, it becomes 110001010.00001111.00010110.
  10100000.
• The first three bits in the last octet, 101,
  indicate that this is the sixth subnetwork. As
  before the remaining bits are all binary 0s. This
  means that the IP address197.15.22.160 must be
  one that is reserved for a subnetwork address.

38
Subnet mask

• How do we determine how many bits to borrow for
  a subnet?
• Determine the number of sub networks required
• Work from the MOST significant (LHS) bits of the
  first octet after the network number and
  calculate the number of bits needed to create the
  required number of subnetworks

39
Subnet mask

• Example
• You are given a class B address and you are
  required to create 1000 subnetworks.
• By working from the LHS of the (first octet after
  the network number) 3rd octet, calculate the
  number of bits to equal or slightly exceed 1000.
  (ie 2x gt 1000)
• This would equate to 210 or 1024-2 networks
• Hence you will need to borrow 10 bits from the
  host portion to create 1000 subnetworks

40
Subnet mask

• Example
• The subnetmask in this instance would
  be255.255.255.192
• How many host per network can you obtain from
  this addressing scheme?

6 bits left for hosts
10 Bits Borrowed (subnetmask)
Natural Class B netmask
41
Subnet mask

• How do we determine how many bits to borrow for
  a subnet given the number of hosts required?
• Determine the number of hosts required
• Work from the LEAST significant (RHS) bits of the
  last octet and calculate the number of bits
  needed to create the required number of
  subnetworks

42
Subnet mask

• Example
• You are given a class B address and you require
  1000 nodes per subnet
• By working from the RHS (last octet) of the 4th
  octet, calculate the number of bits to equal or
  slightly exceed 1000. (ie 2x gt 1000)
• This would equate to 210 or 1024-2 networks
• Hence you will need to borrow 6 bits from the
  host portion to create subnetworks with 1000
  hosts each

43
Subnet mask

• Example
• The subnetmask in this instance would
  be255.255.252.0
• How many subnetworks per network can you obtain
  from this addressing scheme?
• Note Do you recognise this address as the
  student supernet address?

10 bits required for hosts
6 Bits Borrowed (subnetmask)
Natural Class B netmask
44
What about a Supernet?

• A supernet borrows bits from the network
  portion to create contiguous nodes to form a
  super network
• For example
• Company A has about 1000 nodes to address. A
  class B address would be too big (or may not be
  available). Solution Supernetting using 4
  contiguous class C addresses203.10.112.0203.10.
  113.0203.10.114.0203.10.115.0
• (All netmasked to 255.255.255.0)

45
What about a Supernet?

• By allowing the first 2 bits of the third octet
  to be borrowed, a virtual class B address can
  be created.
• A supernet address of 203.10.112.0
  203.10.115.255 is formed with a subnet mask of
  255.255.252.0.
• The Host portion will be expanded from 8 10
  bits
• Route summarisation can occur to 203.10.112.0/22

46
What about a Supernet?
Host portion
Network Portion
203.10.112.0
203.10.113.0
203.10.114.0
203.10.115.255
255.255.252.0
We have expanded the host portion by 2 bits to 10
bits
47
What is a Broadcast Address

• A broadcast address is used to by a node to
  communicate with ALL nodes in a broadcast domain
• Like the netmask, the broadcast address is AND
  with the network address.
• However, the host portion of the network is
  identified in a broadcast address

48
What is a Broadcast Address

• A broadcast address does this by inserting all
  1s in the host portion.
• Eg A natural class B broadcast address would look
  something like this
• N.N.255.255
• If it is not a classful subnetmask, you can
  determine the broadcast address within each
  subnet by locating the host portion and setting
  them to all 1s.

49
What is a Broadcast Address

• An example of a Broadcast address

50
What is a Broadcast Address

• An example of a cross boundary subnet broadcast
  address with a mask of 255.255.252.0

Host portion
Network Portion
SN
Network Address
Host
Broadcast Address
51
What is a Broadcast Address

• An example of a cross boundary subnet broadcast
  address with a mask of 255.255.252.0
• In this example, IP addresses
• 131.181.112.0 131.181.115.255 belong to the
  same subnetwork

Host portion
Network Portion
SN
131.181.112.0(Network)
131.181.113.0
131.181.114.0
131.181.115.255(Broadacast)
52
Network Address VS Broadcast address

• Remember
• A Network address has all the host bits set to
  0
• A Broadcast address has all the host bits set to
  1
• Therefore
• 131.181.112.0 is the network address
• 131.181.115.255 is the broadcast address

53
Network Address VS Broadcast address

• This is important when you are doing ifconfig and
  routing commands
• For example, if a host has an address
  131.181.114.10/22
• The ifconfig route commands would be
• ifconfig ethltxgt inet 131.181.114.10 netmask
  255.255.252.0 broadcast 131.181.115.255
• route add net 131.181.112.0 netmask
  255.255.252.0 dev ethltxgt

54
How do you determine Network and Broadcast
address quickly?

• There are different subnetting exercises
• Given an IP address mask,
• What is the network/subnetwork address
• What is the network/subnetwork broadcast address
• What are the assignable address in that
  network/subnetwork
• What are all the valid subnet addresses
• How many nodes per subnet

55
Algorithm for deriving Subnet information

• Given an IP address, you will usually be given a
  net/subnetmask
• If you are given the mask
• Subtract the mask from 256
• This is known as the multiplier
• The first number in each multiplier value is the
  network number
• The broadcast address is the next multiplier
  value subtract 1

56
Algorithm for deriving Subnet information

• Eg given the IP address 192.168.0.100 with the
  subnet mask of 255.255.255.240 Or
  192.168.0.100/28
• What is the network number
• What is the broadcast address
• What are the valid IP hosts for the subnet

57
Algorithm for deriving Subnet information

• Subtract 256 from the netmask
• 256 - 240 16
• This is the multiplier ie the networks are in
  steps of 16 (16,32,48,64,96,112 etc)
• The IP address 192.168.0.100 is in the range of
  100 / 16 which is the 6 th subnetwork
• The network address is 16 6 96 (01100000b)
• The Broadcast address is 96 16 - 1 gt
  192.168.0.111 (01101111b)
• ie (next multiplier 1)

Host portion
Subnetwork portion
58
Algorithm for deriving Subnet information
Host portion
Network Portion
SN
IP address192.168.0.100
Netmask 255.255.255.240
Network Address192.168.0.96
Broadcast Address192.168.0.111
59
Algorithm for deriving Subnet information

• Valid ranges are
• 192.168.0.97 to 192.168.0.110
• Number of allowable hosts 97 to 110 (incl) 14
  or 2416 - 2 14
• Remember you cannot use the first address
  (network address) and the last address (broadcast
  address) in the range
• The number of allowable networks
• 2416 - 2 14 ( ie 4 bits used. If a class B
  address with the last bit subnet, then add
  another 8 bits to give you 212 2 allowable
  subnet)

60
Algorithm for deriving Subnet information

• What if the IP range goes over 2 octets
• Use the same principal
• Remember octets with all 0s are considered
  boring and will be assigned the mask of 0
• You will then have to locate the position in the
  address with both 1s and 0s (interesting byte)
  and use the same algorithm
• Similarly all 1s are also considered boring and
  will be given the mask of 255 (eg subnetting the
  last byte of a class B address)

61
Algorithm for deriving Subnet information

• Example
• QUT students supernet address
• 131.181.112.0/22
• Netmask expanded 255.255.252.0
• Last byte is not interesting hence we set it to
  0 for network and 1 for broadcast
• The third byte is interesting 256 252 4
  (multiplier)
• Networks are in increments of 4 steps
• 112/4 28 (the 28th subnetwork). Since there is
  no remainder, it is the beginning of the network
  address

62
Algorithm for deriving Subnet information

• Assignable addresses in this subnetwork131.181.11
  2.1 131.181.115.254
• Remember, the 1st and last addresses cannot be
  used (Network and broadcast)

• Network address 131.181.112.0
• Netmask 255.255.252.0
• Broadcast 112 4 1 115 gt131.181.115.255
• Number of valid hosts 210 2 1024 2 1022
  hosts
• Number of subnetworks available for this
  network26 2 64 2 62 subnetworks

63
How does the router handle IP addresses and
subnet masks?

• Let's assume that a device on another network
  with an IP address of 197.15.22.44 wants to send
  data to another device attached to Cisco's
  network with an IP address of 131.108.2.2.
• The data is sent out over the Internet until it
  reaches the router that is attached to Cisco's
  network.
• The router's job is to determine which one of
  Cisco's subnetworks the data should be routed to.
  

64
(No Transcript)
65
when the router performs this ANDing" operation,
the host portion falls through.
66
The router looks at what is left which is the
network number including the subnetwork.
67
The router then looks in its routing table and
tries to match the network number including the
subnet with an interface.
68
How does the "Anding operation change with
different subnet masks?

• Imagine that you have a class "B network with
  the network number 172.16.0.0.
• After assessing the needs of his network, the
  network administrator has decided to borrow eight
  bits in order to create subnetworks.
• When eight bits are borrowed to create subnets,
  the subnet mask is 255.255.255.0.
• Someone outside the network sends data to the IP
  address 172.16.2.120.

69
(No Transcript)
70
(No Transcript)
71
Exercise

• Exercise
• You are given an IP address for a
  host172.168.35.10/20
• What is/are the
• 1. Subnet address?
• 2. Broadcast address?
• 3. The number of useable hosts available for
  this subnet?
• 4. The number of useable subnets available
  for this network?
• 5. The assignable address range for this
  subnet?
• Answers

72
Exercise

• Exercise
• Your organisation has been assigned a class B IP
  address of 130.10.0.0
• You require about 2000 subnetworks
• Work out the
• 1. Subnet mask required for this subnet
• 2. The network and broadcast addresses for
  the first 5 useable subnets
• 3. The number of hosts for each subnet
• 4. The assignable address range of the first
  5 useable subnets
• Answers

73
ROUTING
74
Routing

• Routing describes a process where packets are
  forwarded from one network to another.
• Routing can be performed by devices such as
• dedicated routers
• servers with more than one network interface
  multihomed hosts
• switches incorporating a route function.

75
Routers

• Routing devices typically have more than one
  network interface, each called a port.
• Routers process datagrams individually, making
  routing a processing-intensive operation.
• Dedicated routers offer better performance
  characteristics compared with multi-homed hosts.

76
Routing Tables

• To determining the proper destination network for
  datagrams, routers consult an internal table.
• The table consists of records, one per line, each
  representing a known network.
• Each record includes a set of associated
  characteristics such as netmask

77
Building Route Tables.

• Routing table entries can be built by two
  methods
• Static entries are entered manually by a network
  administrator
• Dynamic entries are entered dynamically by
  routing protocols. Routers learn destination
  network addresses by the periodic exchange of
  route tables between routing devices. Routing
  protocols use IP to deliver this information.

78
Configuring Routers

• Before IP routers can perform the route function
  and use routing protocols to exchange route
  tables, each interface (port) must be correctly
  numbered with a valid host IP address and
  netmask.
• The IP address must be selected from within the
  range for the particular network address.
  Typically local gateways are located the first
  address in the valid host range.

79
Routing Protocols

• Routing protocols are used by routers to
• learn the location of destination networks.
• determine the best route to reach networks.
• Examples of routing protocols include
• RIP, Routing Information Protocol
• OSPF, Open Shortest Path First
• BGP, Border Gateway Protocol

80
Routing Protocols - 2

• Routing protocols differ in
• the way in which they exchange route tables
• determine the route to the destination
• the information that is communicated

81
Distance Vector

• Distance Vector routing protocols broadcast the
  entire route table on a regular basis. RIP2
  typically defaults at once every 30 seconds.This
  creates considerable network traffic.
• They determine the best route path on the basis
  of the least number number of hops to reach a
  destination network.

82
Link State

• Link state protocols only broadcast changes to
  route information after an initial entire table
  has been sent.
• When determining the best path, other factors
  such as policies (e.g. preferred path) and cost
  ( time taken, available bandwidth) can influence
  the choice when multiple paths are available.

83
IP Routing

• Before a routing device can forward an IP
  datagram it must
• examine the Destination Address in the datagram
• use the netmask to identify the network portion
  of the packets destination address
• find a corresponding network address in the route
  table and forward the packet to the gateway or
  interface specified

84
Route Table Fields

• To forward IP datagrams, the router uses the
  following fields of the the route table
• Destination
• Network Mask
• Gateway

85
Destination

• This field lists the networks which are known to
  the router. Addresses may have been entered by an
  administrator, or dynamically learned from the
  transmissions of other routers.
• Address entries concerned with routing between
  network addresses will be of the format
  ltnetidgt, 0

86
Netmask

• The netmask field provides the router with the
  ability to determine the network address of
  packets being examined.
• A logical AND is performed using the netmask and
  the destination address. This logically removes
  the host portion allowing the router to identify
  the destination

87
Gateway

• The Gateway field lists the IP addresses of the
  interface where the datagram should be sent
  (forwarded) to reach the specified Destination.
• This field may contain
• An IP interface address corresponding to an
  adjacent router
• 0.0.0.0
• The address of a interface

88
Destination Hop0.0.0.0

• A Destination of 0.0.0.0 indicates a directly
  connected network. Hosts located on this network
  can be reached using the local network method. If
  the network is Ethernet, the ARP protocol is used
  to find the physical address of the node.