DSL LAN Extender 200 User’s Guide
© 2002 Black Box Corporation
Page 31
Default Behavior
Since all the three parameters are optional, when all of them are turned off, the device is a simple IP
router that works in the following way:
- Configured as CP, it serves as the default gateway for all local machines connected with Ethernet. All
the machines configured this way will send packets with destination address outside the local subnet to
this device, which in turn, sends them to the CO device via the DSL link.
- Configured as CO, it forwards packets coming from CP to its default gateway, or to devices on its local
subnet. Any unicast IP received from the Ethernet will be forwarded to the CP, which in turn examines its
ARP table and forwards the packet to the target end station.
On the CO side, a higher-level router typically learns this route by either static route or by RIP. In the
latter case, the CO’s RIP should be enabled.
If customer implements this network for Internet access, all the PCs served by the CP device must have
public IP addresses (normally assigned by your ISP).
When to Enable Each Option
With the understanding of behavior when all three options are disabled, we can look at each option and
easily tell when to enable them.
RIP enabled:
Enable this option if the device’s Ethernet interface has one or more IP Routers, including other DSL
LAN Extenders.
This parameter should also be enabled on the CO side if it is enabled on the CP side.
This option can also be turned on even if there is no IP router on the local subnet.
NAT Enabled:
Enable this option if you will use the device’s Network Address Translation (NAT) feature. This feature
allows a number of machines to use private IP addresses, and to share a single public IP address when
talking to the outside world.
When NAT is enabled, users actually assign two IP subnets to the CP device’s Ethernet port. CP will
route packets among these two subnets and the outside world (via CO).
The NAT option is not allowed on the CO side.
Bridging Enabled:
There are generally two cases where bridging will be enabled.
NOTE:
The bridging option should be
enabled for both sides if it is enabled on either side.
One is to allow other protocols (like NetWare, NetBEUI, or AppleTalk) to pass through (be bridged)
when IP protocol is routed. In this case, RIP should be enabled even though there is no IP router on the
local Ethernet network. Otherwise, the device can’t decide whether to route or bridge IP packets.
The other is to let all protocols pass through, which makes the two networks connected by the CO-CP pair
a “flat” network. This may sometimes be an easy choice if the remote side (CP side) is a small group
(with a dozen or so or even less machines).
In bridge mode, packets between local machines will be filtered out, but broadcast and multicast packets
are always forwarded to the other side.
