Understanding Internet Protocol 4.p hptx

Understanding
Internet Protocol

Introduction:
As a network administrator, you will use the Transmission Control
Protocol/Internet Protocol (TCP/IP) communications protocol suite most
often. Most techs refer to this simply as Internet Protocol or IP. Although
the newer IPv6 has many advantages over its predecessor, IPv4 is still
used in the majority of local area networks. In this lesson, we will cover
both. To truly be a master of IP networks, a network administrator must
know how the different versions of IP work and how to configure,
analyze, and test them in the GUI and in the command line. By utilizing
knowledge about IP classes and reserved ranges, a well planned
network can be implemented. And by taking advantage of
technologies like network address translation and subnetting, a more
efficient and secure network can be developed. Finally, by incorporating
IPv6 whenever possible, you are opening the door to the future of data
communications and enabling easier administration, bigger and more
powerful data transmissions, and a more secure IP network.

Working with IPv4
Internet Protocol version 4 or IPv4 is the most frequently
used communications protocol. IP resides on the network
layer of the OSI model, and IP addresses consist of four
numbers, each between 0 and 255. The protocol suite is
built into most operating systems and used by most Internet
connections in the United States and many other countries.
As mentioned in Lesson 1, it is composed of a network
portion and a host portion, which are defined by the
subnet mask. In order for an IP address to function, there
must be a properly configured IP address and compatible
subnet mask. To connect to the Internet, you will also need
a gateway address and DNS server address. Advanced
examples of IP configurations include subnetting, network
address translation (NAT), and classless interdomain routing
(CIDR).

Categorizing IPv4 Addresses
 IPv4 addresses have been categorized into five IP
classes. Some have been reserved for private use,
whereas the rest are utilized by public connections. This
classification system helps define what networks can be
used on a LAN and what IP addresses can be used on
the Internet.
 The IPv4 classification system is known as the classful
network architecture and is broken down into five
sections, three of which are commonly used by hosts on
networks—Classes A, B, and C. All five sections are
displayed in Table 4-1. The first octet of the IP address
defines which class the address is a member of.

Table 4.1 IPv4 classful network architecture

Class A network addresses are used by the government, ISPs,
big corporations, and large universities. Class B network
addresses are used by mid-sized companies and smaller ISPs.
Class C network addresses are used by small offices and
home offices.
In the table, the term node is synonymous with “host.” If an IP
address is Class A, the first octet is considered to be the
“network” portion. The other three octets are then the node or
host portion of the address. So, a computer might be on the 11
network and have an individual host ID of 38.250.1, making the
entire IP address 11.38.250.1. In looking at the table, you might
also have noticed a pattern. In particular, Class B addresses use
two octets as the network portion (e.g., 128.1). The other two
octets are the host portion. Meanwhile, Class C addresses use
the first three octets as the network portion (e.g., 192.168.1).
Here, the last octet is the host portion.
There are several other notations we need to make to this table.

First, as shown, the range for Class A is 0–127. However, the 127 network
number isn’t used by hosts as a logical IP address. Instead, this network is
used for loopback IP addresses, which allow for testing. For example, every
computer that runs IPv4 is assigned a logical IP address such as 192.168.1.1.
However, every computer is also automatically assigned the address
127.0.0.1, and any address on the 127 network (for example, 127.200.16.1)
redirects to the local loopback. Therefore, this network number cannot be
used when designing your logical IP network, but it can definitely be used
to aid in testing. Second, as you look at Table 4-1, note the default subnet
masks for each class. Notice how they ascend in a corresponding fashion
to the network/node portions. Memorize the default subnet masks for Class
A, B, and C.

Third, be aware that the total number of usable addresses is always going
to be two less than the mathematical amount. For example, in a Class C
network such as 192.168.50.0, there are 256 mathematical values: the
numbers including and between 0 and 255. However, the first and last
addresses can’t be used. The number 0 and the number 255 cannot be
used as logical IP addresses for hosts because they are already utilized
automatically. The 0 in the last octet of 192.168.50.0 defines a network
number, not a single IP address, it is the entire network. And 192.168.50.255
is known as the broadcast address, which is used to communicate with all
hosts on the network. So, because you can never use the first and last
addresses, you are left with two fewer addresses—in this case, 254 usable
IP addresses. This applies to bigger networks as well. For instance, a Class A
network can use 16,777,214 addresses instead of 16,777,216. If we examine
this more carefully, we will see that the number zero in binary equals
00000000 and the number 255 in binary is 11111111. Thus, we can’t use the
“all zeros” octet and the “all ones” octet. This rule applies to total hosts,
but not to total networks within a particular class. We build on this concept
in the subnetting section later in this lesson.

One other related notion is the network 0, which generally isn’t used but is
listed in the table because it is technically considered part of Class A.
Next, Class D and Class E are not used by regular hosts. Therefore, they
are not given a network/node classification, and as a result of that, they
are not given a specific number of networks or total hosts they can utilize.
Instead, Class D is used for what is known as multicasting—transmitting
data to multiple computers (or routers). Class E was reserved for future
use, but this has given way to IPv6 instead.
Finally, try to get into the habit of converting IP octets into their binary
form. For example, the binary range of the first octet in Class A (0–127) is
00000000–01111111. For Class B, it is 10000000–10111111, and for Class C, it
is 11000000–11011111. To practice doing this, you can use one of many
decimal-to-binary conversion methods (such as the one shown in Table 4-
2), or for now, you can use the scientific calculator in Windows by
navigating to the Run prompt and typing calc.exe. Then click View on the
calculator’s menu bar and select Scientific. This will help you when it
comes to more complex IP networks and when you attempt to create
subnetworks. Keep in mind that computer certification exams might not
allow use of a calculator.

IPv4 addresses are further classified as either public or private.
Public IP addresses are ones that are exposed to the Internet;
any other computers on the Internet can potentially
communicate with them. Private IP addresses are hidden from
the Internet and any other networks. They are usually behind
an IP proxy or firewall device. There are several ranges of
private IP addresses that have been reserved by the IANA, as
shown in Table 4-3. The majority of the other IPv4 addresses
are considered public.
Table 4.3 Private IPv4
addresses as assigned
by IANA

The only private Class A network is 10. However, there are multiple Class B
and C private networks. 172.16, 172.17, and so on through 172.31 are all
valid private Class B networks. And 192.168.0, 192.168.1, 192.168.2, and so
on all the way through 192.168.255 are all valid private Class C networks.
Remember that for an address to be Class C, the first three octets must be
part of the network portion; for Class B, the first and second octets; and for
Class A, only the first octet.
Another type of private range was developed by Microsoft for use on
small peer-to-peer Windows networks. It is called APIPA, which is an
acronym for Automatic Private IP Addressing. It uses a single Class B
network number: 169.254.0.0. If a Windows client cannot get an IP
address from a DHCP server and has not been configured statically, it
will autoassign a number on this network. If, for some reason, APIPA
assigns addresses even though a DHCP server exists, APIPA can be
disabled in the registry. See the Microsoft Support site for details.

Default Gateways and DNS Servers
To complete our IP configuration, we need a default gateway address and a DNS
server address. This will help our client computers access the Internet
The first such field is the default gateway field. The default gateway is the first IP
address of the device that a client computer will look for when attempting to gain
access outside the local network. This device could be a router, server, or other
similar device; it is the device that grants access to the Internet or other networks.
This device’s address is on the same network number as the client. So, for example, if
the client is 192.168.50.1, the gateway might be 192.168.50.100. Many gateway
devices come preconfigured with their own LAN IP, but this is almost always
configurable. For example, the D-Link DIR-655 we accessed in the previous lesson
was configured as 192.168.0.1, but we could change that if we wanted to. Without a
default gateway address configured within our local computer’s IP Properties dialog
box, we cannot gain access to any other networks. It is possible to have more than
one gateway address in case the default gateway device fails. This can be done in
Windows 7 by navigating to the Network Connections window, right clicking the
network adapter in question (for example, Local Area Connection), selecting
Properties, selecting Internet Protocol Version 4, and selecting the Properties button.
In the Internet Protocol Version 4 Properties dialog box, click the Advanced button.
Additional gateway addresses can be added to the Default gateways field.

The second field we need to configure is the DNS server address. The DNS server
address is the IP address of the device or server that resolves DNS addresses to IP
addresses. This could be a Windows Server or an all-in-one multifunction network
device—it depends on the network environment. Also, it could be on the LAN
(common in large networks) or located on the Internet (common in smaller
networks). One example of a name resolution would be the domain name
www.google.com, which currently resolves to the IP address 66.249.91.104. To
demonstrate this, try typing this command in the command prompt: ping
www.google. com. You should get results similar to “Reply from 66.249.91.104…”.
Google can change its IP address at any time, but the results should be similar.
By the way, this is an example of a public IP address. The whole concept here is
that computers ultimately communicate by IP address. However, it is easier for
people to remember www.google.com than it is for them to remember an IP
address. The DNS server resolves domain names like www.proseware.com, host
names like server1.proseware.com, and so on. Without this DNS server address, a
client computer will not be able to connect by name to any resource on the
Internet. DNS servers are also necessary in Microsoft domain environments. If your
computer is a member of such an environment and the DNS server address is
not configured properly, domain resources will most likely be inaccessible.

Defining Advanced IPv4 Concepts
Methods such as network address translation, subnetting,
and classless inter-domain routing (CIDR) can make
networks faster, more efficient, and more secure. These
advanced IP configurations are found in most networks
today. Therefore, to be a proficient network engineer, you
must master these concepts.

NETWORK ADDRESS TRANSLATION
Network address translation (NAT) is the process of modifying an IP
address while it is in transit across a router, computer, or similar device.
This is usually so one larger address space (private) can be re-mapped
to another address space, or perhaps re-mapped to a single public IP
address. This process is also known as IP masquerading, and it was
originally implemented due to the problem of IPv4 address exhaustion.
Today, NAT hides a person’s private internal IP address, making it more
secure. Some routers only allow for basic NAT, which carries out IP
address translation only. However, more advanced routers allow for
port address translation (PAT), a subset of NAT, which translates both IP
addresses and port numbers. A NAT implementation on a firewall hides
an entire private network of IP addresses (e.g., the 192.168.50.0
network) behind a single publicly displayed IP address. Many SOHO
routers, servers, and similar devices offer this technology to protect a
company’s computers on a LAN from outside intrusion.

Figure 4-6 illustrates how NAT might be implemented with some fictitious IP
addresses. Here, the router has two network connections. One goes to the
LAN—192.168.50.254—and is a private IP address. This is also known as an
Ethernet address and is sometimes referred to as E0 or the first Ethernet
address. The other connection goes to the Internet or WAN— 64.51.216.27
—and is a public IP address. Sometimes, this will be referred to as S0 ,
which denotes a serial address (common to vendors such as Cisco). So,
the router is employing NAT to protect all of the organization’s computers
(and switches) on the LAN from possible attacks initiated by mischievous
persons on the Internet or in other locations outside the LAN.
Fig 4.6 NAT implementation

SUBNETTING
Subnetting could be considered one of the most difficult concepts in
networking—but it can be simplified with some easy equations and a
well-planned implementation process. Until now, we have used default
subnet masks. However, one reason for having a subnet mask is to gain
the ability to create subnetworks logically by IP. We must ask, what is a
subnet? It is a subdivision of your logical IP network; by default, all
computers are on one subnet or network with no divisions involved.
And . . . what is a mask? It is any binary number that is a 1. If the binary
digit is a 1, then it is masked. If the binary digit is a 0, then it is
unmasked. Let’s review the standard default subnet masks, as shown in
Table 4-4.
Tab 4.4 Standard subnet mask
review

There are a lot of different subnetting options, but as one example, we
could use 255.255.255.240. This would also be known as 192.168.1.0 /28
because the binary equivalent of the subnet mask has 28 masked bits and
4 unmasked bits.
The first three 255s are the same, and we can pretty much ignore them,
but the fourth octet (240) tells us how many subnetworks (subnet IDs) and
hosts we can have per subnetwork. All you need is the ability to convert to
binary and to use two equations:
 • Equation #1: 2n
= x
 • Equation #2: 2n
– 2 = x
 1. Convert 240 to binary. It equals 11110000.
 2. Break the octet up like this: 1111 and 0000. Use the part made up of 1s for the subnet IDs
and the part made up of 0s for the host IDs.
 3. To find out the total number of subdivisions (or subnet IDs) you can have in your network,
input the amount of 1s into equation #1. There are four 1s in 11110000, so the number 4
should replace n , making the equation 24
= x. Because 24
= 16, this means the maximum
number of subnets is 16. However, it is recommended that the first and last subnets not be
used. That leaves us with 14 usable subnets.
 4. But (and there’s always a but . . .) you can never use the first and the last IP address for a
host ID. “All Ones” and “All Zeros” cannot be used as they are for identifying the
subnetwork and for doing broadcasting. To find out the total number of hosts per subnet
you can use in your network input the number of 0s into equation #2. There just happen to
be four 0s in 11110000. Therefore, the number 4 should replace n , making the equation 24
–
2 = x. Because 24
– 2 = 14, the maximum number of hosts per subnet is 14.

Table 4.5 Possible subnets and hosts in the 192.168.50.0/28 subnet working
scenario

Defining Classless Inter-Domain Routing (CIDR)
 Classless inter-domain routing (CIDR) is a way of allocating IP
addresses and routing Internet Protocol packets. It was intended to
replace the prior classful IP addressing architecture in an attempt to
slow the exhaustion of IPv4 addresses. Classless inter-domain routing
is based on variable-length subnet masking (VLSM), which allows a
network to be divided into different-sized subnets to make one IP
network that would have previously been considered a class (such
as Class A) look like Class B or Class C. This can help network
administrators efficiently use subnets without wasting IP addresses.
One example of CIDR would be the IP network number
192.168.0.0/16. The /16 means that the subnet mask has 16 masked
bits (or 1s) making 255.255.0.0. Usually, that would be a default Class
B subnet mask, but because we are using it in conjunction with
what used to be a Class C network number, the whole kit and
caboodle becomes classless.

Working with IPv6
IPv6 is the new generation of IP addressing for the Internet, but it can also
be used in small office networks and home networks. It was designed to
overcome the limitations of IPv4, including address space and security.
Understanding IPv6
IPv6 has been defined for over a decade, and it has slowly been gaining
acceptance in the networking world, although it is still considered in its
infancy. The number-one reason to use IPv6 is address space. IPv6 is a 128-
bit system, whereas its still-dominant predecessor IPv4 is only a 32-bit
system. What does this mean? Well, whereas IPv4 can have approximately
4 billion IP addresses in the whole system, IPv6 can have 340 undecillion
addresses. That’s 340 with 36 zeroes after it! Of course, various limitations in
the system will reduce that number, but the final result is still far greater
than with the IPv4 system. Yet another reason to use IPv6 is advanced
integrated security; for example, IPSec is a fundamental component of
IPv6 (we will discuss IPSec in more depth in Lesson 6). IPv6 also has many
advancements and simplifications when it comes to address assignment.
Table 4-9 summarizes some of the differences between IPv4 and IPv6.

Table 4.9 IPv4 versus IPv6
IPv6 also supports jumbograms. These are much larger packets than IPv4 can
handle. IPv4 packets are normally around 1,500 bytes in size, but they can go as
large as 65,535 bytes. In comparison, IPv6 packets can optionally be as big as
approximately 4 billion bytes. We mentioned already that IPv6 addresses are
128-bit numbers. They are also hexadecimal in format and divided into eight
groups of four numbers each, with each group separated by a colon. These
colon separators contrast with IPv4’s dot-decimal notation. In Windows, IPv6
addresses are automatically assigned and auto-configured, and they are known
as link local addresses. There are three main types of IPv6 addresses:

Types of IPV6
 • Unicast address: This is a single address on a single interface. There
are two types of unicast addresses. The first, global unicast addresses,
are routable and displayed directly to the Internet. These addresses
start at the 2000 range. The other type is the aforementioned link local
address. These are further broken down into two subtypes, the Windows
auto-configured address, which starts at either FE80, FE90, FEA0 and
FEB0, and the loopback address, which is known as ::1, where ::1 is the
equivalent of IPv4’s 127.0.0.1.
 • Anycast address: These are addresses assigned to a group of
interfaces, most likely on separate hosts. Packets that are sent to these
addresses are delivered to only one of the interfaces—generally, the
first one, or closest, available. These addresses are used in failover
systems.
 • Multicast address: These addresses are also assigned to a group of
interfaces and are also most likely on separate hosts, but packets sent
to such an address are delivered to all of the interfaces in the group.
This is similar to IPv4 broadcast addresses (such as 192.168.1.255).
Multicast addresses do not suffer from broadcast storms the way their
IPv4 counterparts do.

DEFINING THE DUAL IP STACK
 A dual IP stack exists when there are two Internet Protocol software
implementations in an operating system, one for IPv4 and another
for IPv6. Dual stack IP hosts can run IPv4 and IPv6 independently, or
they can use a hybrid implementation, which is the most commonly
used method for modern operating systems.
 Dual stack TCP/IP implementations enable programmers to write
networking code that works transparently on IPv4 or IPv6. The
software can use hybrid sockets designed to accept both IPv4 and
IPv6 packets. When used in IPv4 communications, hybrid stacks use
IPv6 methodologies but represent IPv4 addresses in a special IPv6
address format known as the IPv4-mapped address. IPv4-mapped
addresses have the first 80 bits set to 0 (note the double colon), the
next 16 set to 1 (shown as ffff), and the last 32 bits populated by the
IPv4 address. These addresses look like IPv6 addresses, other than the
last 32 bits, which are written in the customary dotdecimal notation.
Here is an example:
 ::ffff:10.254.254.1
 This is an IPv4-mapped IPv6 address for the IPv4 address 10.254.254.1

DEFINING IPv4 TO IPv6 TUNNELING
 IPv6 packets can be encapsulated inside IPv4 datagrams. This is
known as IPv6 tunneling, or IP 6 to 4. In Microsoft operating systems,
this is generally done with the Teredo adapter, which is a virtual
adapter or “pseudo-interface,” not a physical network adapter. This
allows connectivity for IPv6 hosts that are behind an IPv4 device or
IPv6 unaware device. It ensures backward compatibility. An example
of one of these addresses would be:
 Fe80::5efe:10.0.0.2%2
 Notice that this is a link-local address and that the IPv4 address
(10.0.0.2) is actually part of the whole IPv6 address. IPv6 tunneling
requires little router configuration and no client computer
configuration whatsoever, so it is fairly easy to implement, enabling
IPv6 clients to interact with IPv6 servers on the Internet, even though
the router is not IPv6 aware.

Understanding Internet Protocol 4.p hptx

More Related Content

Similar to Understanding Internet Protocol 4.p hptx (20)

Recently uploaded (20)

Understanding Internet Protocol 4.p hptx