System information

Appendix A: Advanced Device Configuration 375
Address Configuration for Connected Devices
For recommendations about which ranges to use for various sizes of
organizations and for avoiding address conflicts, see http://
www.rhebus.com/techinfo/iprange.htm#ip1.
The number of IP address available on a network may be restricted by a
subnet mask. For a simple example, the subnet mask 255.255.255.0 provides
256 IP addresses. The IP address ending with zero (0) is the network address,
and the IP address ending with 255 is the broadcast address, leaving 254
addresses to assign to devices (from 1-254).
To specify a range of addresses on the AlterPath OnBoard supply the network
address and a subnet mask, in either of these two formats: 192.168.0.0 and
255.255.255.0 or 192.168.0.0/24.
Why Define Private Subnets?
At least one private subnet must be defined on the OnBoard for the following
purposes:
To define a private OnBoard address for the OnBoard and connected
devices to use when communicating.
To enable communications between remote users workstations on the
Internet or local users on the same LAN and connected devices on the
private management network, via the OnBoard’s Native IP access facility.
The private Ethernet ports are accessed through the priv0 interface on the
OnBoard, which interacts with connected devices through an internal switch.
The OnBoard attempts to reach a device that does not have a private subnet
assigned by attempting to contact it through the OnBoard’s default route.
Therefore, unless the OnBoard administrator defines a private subnet and
assigns it to each device, the device cannot be reached unless the device is on
the public side of the OnBoard. In almost all cases, devices are on the private
side of the OnBoard and therefore they are unreachable without a private
subnet.
The following should be kept in mind when planning the addressing scheme:
When the connected devices’ addresses are all within the same range,
only one private subnet is required.
The administrator should assign IP addresses to all service processors
from the same block of addresses, if possible, to make it possible to
administer the IP addresses using only a single private subnet.