Windows Server 2012 R2 Inside Out: Networking with TCP/IP

Using subnets and subnet masks

Anyone who works with computers should learn about subnetting and what it means. A subnet is a portion of a network that operates as a separate network. Logically, it exists separately from other networks, even if hosts on those other networks share the same network ID. Typically, such networks are also physically separated by a router. This ensures that the subnet is isolated and doesn’t affect other subnets.

Subnetting is designed to make more efficient use of the IPv4 address space. Thus, rather than having networks with hundreds, thousands, or millions of nodes, you have a subnet that is sized appropriately for the number of nodes you use. This is important, especially for the crowded public IPv4 address space, where it doesn’t make sense to assign the complete IPv4 address range for a network to an individual organization. Thus, instead of getting a complete network address for the public Internet, your organization is more likely to get a block of consecutive IPv4 addresses to use.

Subnet masks

You use a 32-bit value known as a subnet mask to configure nodes in a subnet to communicate only with other nodes on the same subnet. The mask works by blocking areas outside the subnet so that they aren’t visible from within the subnet. Because they are 32-bit values, subnet masks can be expressed as an address for which each 8-bit value (octet) is written as four separate decimal values delimited by a period (dot). As with IPv4 addresses, the basic form is w.x.y.z.

The subnet mask identifies which bits of the IPv4 address belong to the network ID and which bits belong to the host ID. Nodes can see only the portions of the IPv4 address space that aren’t masked by a bit with a value of 1. If a bit is set to 1, it corresponds to a bit in the network ID. If a bit is set to 0, it corresponds to a bit in the host ID.

Because a subnet mask must be configured for each IPv4 address, nodes on both classful and nonclassful networks have subnet masks. On a classful network, all the bits in the network ID portion of the IPv4 address are set to 1 and can be presented in dotted decimal, as shown in Table 2-5.

Table 2-5 Standard subnet masks for classful networks

Address Class	Bits for Subnet Mask	Subnet Mask
Class A	11111111 00000000 00000000 00000000	255.0.0.0
Class B	11111111 11111111 00000000 00000000	255.255.0.0
Class C	11111111 11111111 11111111 00000000	255.255.255.0

Network prefix notation

With subnetting, an IPv4 address alone doesn’t help you understand how the address can be used. To be sure, you must know the number of bits in the network ID. As discussed, the subnet mask provides one way to determine which bits in the IPv4 address belong to the network ID and which bits belong to the host ID. If you have a block of IPv4 addresses, writing out each IPv4 address and the subnet mask is rather tedious. A shorthand way to do this is to use network prefix notation, which is also referred to as the classless inter-domain routing (CIDR) notation.

In network prefix notation, the network ID is seen as the prefix of an IPv4 address and the host ID is seen as the suffix. To write a block of IPv4 addresses and specify which bits are used for the network ID, you write the network number followed by a forward slash and the number of bits in the network ID, as in the following:

NetworkNumber/# of bits in the network ID

The slash and the number of bits in the network ID are referred to as the network prefix. Following this, you could rewrite Table 2-5 in the way shown in Table 2-6.

Table 2-6 Standard network prefixes for classful networks

Address Class	Bits for Subnet Mask	Network Prefix
Class A	11111111 00000000 00000000 00000000	/8
Class B	11111111 11111111 00000000 00000000	/16
Class C	11111111 11111111 11111111 00000000	/24

You now have two ways of detailing which bits are used for the network ID and which bits are used for the host ID. With the network number 192.168.1.0, you could use either of the following to specify that the first 24 bits identify the network ID:

• 192.168.1.0, 255.255.255.0
• 192.168.1.0/24

With either entry, you know that the first 24 bits identify the network ID and the last 8 bits identify the host ID. This, in turn, means that the usable IPv4 addresses are 192.168.1.1 through 192.168.1.254.

Subnetting

When you use subnetting, nodes no longer follow the class rules for determining which bits in the IPv4 address are used for the network ID and which bits are used for the host ID. Instead, you set the 32 bits of the IPv4 address as appropriate to be either network ID bits or host ID bits based on the number of subnets you need and then number nodes for each subnet. There is an inverse relationship between the number of subnets and the number of nodes per subnet that can be supported. As the number of subnets goes up by a factor of two, the number of hosts per subnet goes down by a factor of two.

Because Class A, B, and C networks have a different number of host ID bits to start with, borrowing bits from the host ID yields different numbers of subnets and hosts. The technique is the same, however. Each bit represented as a 1 in the subnet mask corresponds to a bit that belongs to the network ID. This means the value of each bit can be represented as shown in Figure 2-6.

You start with the high-order bits and work your way to the low-order bits. When you borrow 1 bit of the host ID, you increase the number of possible subnets by a factor of two and reduce the number of possible hosts by a factor of two.

Subnetting Class A networks

The network entry mask for a standard Class A network can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class A	11111111 00000000 00000000 00000000	/8	255.0.0.0

If you want to divide a Class A network into two separate subnets, you can borrow the high-order bit from the host ID in the second octet and add this bit to the network ID. Because the value of this bit taken from the host ID is 128, the corresponding subnet mask is 255.128.0.0. Thus, the network entry for the subnetted Class A network can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class A	11111111 10000000 00000000 00000000	/9	255.128.0.0

If you take an additional bit from the host ID bits, you allow the Class A network to be divided into up to four subnets. The value of this bit taken from the host ID is 64. When you add this value to the value of the previous bit taken from the host ID, the sum is 192 (128 + 64) and the corresponding subnet mask is 255.192.0.0. This means that the network entry for a subnetted Class A network that can be divided into up to four subnets can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class A	11111111 11000000 00000000 00000000	/10	255.192.0.0

Table 2-7 shows how Class A networks can be subnetted and how this affects the number of possible subnets and hosts per subnet.

Table 2-7 Subnetting Class A networks

Maximum Subnets	Bits for Subnet Mask	Network Prefix	Decimal	Maximum Nodes
1	11111111 00000000 00000000 00000000	/8	255.0.0.0	16,777,214
2	11111111 10000000 00000000 00000000	/9	255.128.0.0	8,388,606
4	11111111 11000000 00000000 00000000	/10	255.192.0.0	4,194,302
8	11111111 11100000 00000000 00000000	/11	255.224.0.0	2,097,150
16	11111111 11110000 00000000 00000000	/12	255.240.0.0	1,048,574
32	11111111 11111000 00000000 00000000	/13	255.248.0.0	524,286
64	11111111 11111100 00000000 00000000	/14	255.252.0.0	262,142
128	11111111 11111110 00000000 00000000	/15	255.254.0.0	131,070
256	11111111 11111111 00000000 00000000	/16	255.255.0.0	65,534
512	11111111 11111111 10000000 00000000	/17	255.255.128.0	32,766
1024	11111111 11111111 11000000 00000000	/18	255.255.192.0	16,382
2048	11111111 11111111 11100000 00000000	/19	255.255.224.0	8190
4096	11111111 11111111 11110000 00000000	/20	255.255.240.0	4094
8192	11111111 11111111 11111000 00000000	/21	255.255.248.0	2046
16,384	11111111 11111111 11111100 00000000	/22	255.255.252.0	1022
32,768	11111111 11111111 11111110 00000000	/23	255.255.254.0	510
65,536	11111111 11111111 11111111 00000000	/24	255.255.255.0	254
131,072	11111111 11111111 11111111 10000000	/25	255.255.255.128	126
262,144	11111111 11111111 11111111 11000000	/26	255.255.255.192	62
524,288	11111111 11111111 11111111 11100000	/27	255.255.255.224	30
1,048,576	11111111 11111111 11111111 11110000	/28	255.255.255.240	14
2,097,152	11111111 11111111 11111111 11111000	/29	255.255.255.248	6
4,194,304	11111111 11111111 11111111 11111100	/30	255.255.255.252	2

Subnetting Class B networks

The network entry mask for a standard Class B network can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class B	11111111 11111111 00000000 00000000	/16	255.255.0.0

A standard Class B network can have up to 65,534 hosts. If you want to divide a Class B network into two separate subnets, you can borrow the high-order bit from the host ID in the third octet and add this bit to the network ID. Because the value of this bit taken from the host ID is 128, the corresponding subnet mask is 255.255.128.0. Thus, the network entry for the subnetted Class B network can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class B	11111111 11111111 10000000 00000000	/17	255.255.128. 0

If you take an additional bit from the host ID bits, you allow the Class B network to be divided into up to four subnets. The value of this bit taken from the host ID is 64. When you add this value to the value of the previous bit taken from the host ID, the sum is 192 (128 + 64) and the corresponding subnet mask is 255.255.192.0. This means that the network entry for a subnetted Class B network that can be divided into up to four subnets can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class B	11111111 11111111 11000000 00000000	/18	255.255.192.0

Table 2-8 shows how Class B networks can be subnetted and how this affects the number of possible subnets and hosts per subnet.

Table 2-8 Subnetting Class B networks

Maximum Subnets	Bits for Subnet Mask	Network Prefix	Decimal	Maximum Nodes
1	11111111 11111111 00000000 00000000	/16	255.255.0.0	65,534
2	11111111 11111111 10000000 00000000	/17	255.255.128.0	32,766
4	11111111 11111111 11000000 00000000	/18	255.255.192.0	16,382
8	11111111 11111111 11100000 00000000	/19	255.255.224.0	8190
16	11111111 11111111 11110000 00000000	/20	255.255.240.0	4094
32	11111111 11111111 11111000 00000000	/21	255.255.248.0	2046
64	11111111 11111111 11111100 00000000	/22	255.255.252.0	1,022
128	11111111 11111111 11111110 00000000	/23	255.255.254.0	510
256	11111111 11111111 11111111 00000000	/24	255.255.255.0	254
512	11111111 11111111 11111111 10000000	/25	255.255.255.128	126
1024	11111111 11111111 11111111 11000000	/26	255.255.255.192	62
2048	11111111 11111111 11111111 11100000	/27	255.255.255.224	30
4096	11111111 11111111 11111111 11110000	/28	255.255.255.240	14
8192	11111111 11111111 11111111 11111000	/29	255.255.255.248	6
16,384	11111111 11111111 11111111 11111100	/30	255.255.255.252	2

Subnetting Class C networks

The network entry mask for a standard Class C network can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class C	11111111 11111111 11111111 00000000	/24	255.255.255.0

A standard Class C network can have up to 254 hosts. If you want to divide a Class C network into two separate subnets, you can borrow the high-order bit from the host ID in the fourth octet and add this bit to the network ID. Because the value of this bit taken from the host ID is 128, the corresponding subnet mask is 255.255.255.128. Thus, the network entry for the subnetted Class C network can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class C	11111111 11111111 11111111 10000000	/25	255.255.255.128

If you take an additional bit from the host ID bits, you allow the Class C network to be divided into up to four subnets. The value of this bit taken from the host ID is 64. When you add this value to the value of the previous bit taken from the host ID, the sum is 192 (128 + 64) and the corresponding subnet mask is 255.255.255.192. This means that the network entry for a subnetted Class C network that can be divided into up to four subnets can be defined as follows:

Address Class	Bits for Subnet Mask	Network Prefix	Decimal
Class C	11111111 11111111 11111111 11000000	/26	255.255.255.192

Table 2-9 shows how Class C networks can be subnetted and how this affects the number of possible subnets and hosts per subnet.

Table 2-9 Subnetting Class C networks

Maximum Subnets	Bits for Subnet Mask	Network Prefix	Decimal	Maximum Nodes
1	11111111 11111111 11111111 00000000	/24	255.255.255.0	254
2	11111111 11111111 11111111 10000000	/25	255.255.255.128	126
4	11111111 11111111 11111111 11000000	/26	255.255.255.192	62
8	11111111 11111111 11111111 11100000	/27	255.255.255.224	30
16	11111111 11111111 11111111 11110000	/28	255.255.255.240	14
32	11111111 11111111 11111111 11111000	/29	255.255.255.248	6
64	11111111 11111111 11111111 11111100	/30	255.255.255.252	2

Understanding IP data packets

With IPv4, computers send data in discrete packets of information with a header and a payload. IPv4 headers are variable in size, between 20 and 60 bytes, in 4-byte increments. Each bit range is broken into different sections, and each section corresponds to the range of a related field in a packet. Header bit ranges consist of 0–3, 4–7, 8–15, 16–18, and 15–31. These correspond to the values 0, 32, 64, 96, 128, 160, and 160/152+ for data.

For examples of the ranges and their use, see Table 2-10. The IP payload is of variable size as well, ranging from 8 bytes to 65,515 bytes. Although most people will never use this information on a regular basis, it’s very useful for understanding how to troubleshoot network problems.

Table 2-10 IPv4 packets

+	Bits 0-3	4-7	8-15	16-18	15-31
0	Version	Header length	Type of service	Total length
32	Identification			Flags	Fragment offset
64	Time to Live (TTL)		Protocol	Header checksum
96	Source address information
128	Destination address information
160	Optional information
160/152+	Data transmitted