System information

142
AS/400 IBM Network Station: Techniques for Deployment in a WAN
Looking at the Figure 49 on page 141, a mask of 128 in the last octet provides
two address groups:
• .1 to .126
• .129 to .254
The subnet boundaries in both groups cannot be used. In this case, the
addresses .0, .127, .128 and .255 cannot be used.
The same applies for a mask of 240. This mask gives you 16 groups of (16-2)
contiguous addresses, remembering that the boundary addresses in each
group, cannot be used.
For selecting twinax subnets,
only consider mask settings of 248, 240, 224
and 192 because these subnets provide a number of IP addresses that are
supported by the workstation controller. A mask of 128 or 0 can be assigned.
However, many addresses in these groups cannot be used and hence are
wasted. As previously mentioned, detailed tables outlining the combinations
of subnet masks and groups of IP addresses can be found in
IBM Network
Station Manager Installation and Use,
SC41-0664 .
The next example uses a class
C
TCP/IP address that is divided into four
different address groups. Three subnet groups, for three different TCP/IP over
twinax networks, is assigned. A large group of addresses are allocated for the
rest of the network.
Figure 50 on page 143 shows the Network Topology for this example and
Figure 51 on page 143 shows the division of the address space 192.168.1.x.
The following example does not follow the recommendation of allowing a
maximum contiguous range of 64 TCP/IP addresses allocated to the twinax
subnet. It is only intended to provide an example of transparent subnetting
in a twinax environment. Unless you are limited, by your own network IP
addressing scheme, to using such an example, allocate the maximum
number of IP addresses to the twinax subnet.
Note