z/OS Communications Server: IP Diagnosis Guide
Previous topic | Next topic | Contents | Contact z/OS | Library | PDF

IPv4 subnetting

z/OS Communications Server: IP Diagnosis Guide

IPv4 allows for a variation of the network and host segments of an IP address, known as subnetting, can be used to physically and logically design a network. For example, an organization can have a single internet network address (NETID) that is known to users outside the organization, yet configure its internal network into different departmental subnets. Subnetwork addresses enhance local routing capabilities, while reducing the number of network addresses required.

To illustrate this, consider a simple example. Assume that we have an assigned class C network address of 192.9.200 for our site. This would mean that we could have host addresses from to If we did not use subnetting, then we could only implement a single IP network with 254 hosts. To split our site into two logical subnetworks, we could implement the network scheme shown in Figure 1:
Figure 1. Subnetting scheme
Without Subnetting:
                                          Network        Host Address
                                          Address        Range
       192      9        200       host
    11000000 00001001 11001000 xxxxxxxx   192.9.200      1 - 254
With Subnetting:
                                          Subnet         Host Address  Subnet
                                          Address        Range         Value
       192      9        200   64  host
    11000000 00001001 11001000 01xxxxxx   65 - 126      01
                                          Subnet         Host Address  Subnet
                                          Address        Range         Value
       192      9        200   128 host
    11000000 00001001 11001000 10xxxxxx  129 - 190     10
The subnet mask would be
       255      255      255     192
    11111111 11111111 11111111 11000000
z/OS® TCP/IP uses a slightly different scheme for the subnet mask when defining the BEGINROUTES statements in the hlq.PROFILE.TCPIP data set and for displaying the subnet mask within a Netstat GATE/-g command. The subnet mask is applied only to the host segment of the IP address, and Netstat GATE/-g displays the subnet mask for only the host segment of the IP address. The subnet mask in the preceding chart as defined for z/OS TCP/IP would be:
   0        0        0       192
00000000 00000000 00000000 11000000

Although z/OS TCP/IP defines the subnet mask differently, the application of the subnet mask and subnet value to the IP address is consistent with RFC-architected routing algorithms. A subnet mask of 255 is used for the remainder of this section, to retain symmetry with other routing documents that use 255 as the subnet value for the network segment of an IP address.

Because subnets B'00' and B'11' are both reserved, only two subnets are available. All 0s and all 1s have a special significance in internet addressing and should be used with care. Also notice that the total number of host addresses that we can use is reduced for the same reason. For instance, we cannot have a host address of 16 because this would mean that the subnet/host segment of the address would be B'0001000', which with the subnet mask we are using, would mean a subnet value of B'00', which is reserved.

The same is true for the host segment of the fourth octet. A fourth octet value of B'01111111' is reserved because, although the subnet of B'01' is valid, the host value of B'1' is reserved.

The network segment of the subnet mask is always assumed to be one, so each octet has a decimal value of 255. For example, with a class B address, the first two octets are assumed to be 255.255.

Go to the previous page Go to the next page

Copyright IBM Corporation 1990, 2014