Stub network

From Citizendium
Jump to navigation Jump to search
This article may be deleted soon.
To oppose or discuss a nomination, please go to CZ:Proposed for deletion and follow the instructions.

For the monthly nomination lists, see
Category:Articles for deletion.


This article is developing and not approved.
Main Article
Discussion
Related Articles  [?]
Bibliography  [?]
External Links  [?]
Citable Version  [?]
 
This editable Main Article is under development and subject to a disclaimer.

In Internet Protocol networking, a stub network is a part of the network topology that can be used to reach hosts local to it, not other networks. A very similar concept applies in the Public Switched Telephone Network, where end office switches know how to reach directly connected telephones, and a default on where to go to reach other exchanges, higher level parts of a national telephone number, and international destinations.

Image:Routing-Stub-2LAN.png|thumb|left|500px|Two stub networks The first topology diagram shows a pair of broadcast-capable stub networks (e.g., Ethernet/IEEE 802.3), with Internet Protocol version 4 addressing, interconnected by a single router. Under the local versus remote assumption of IP routing, any host on the 192.168.0.0/24 network knows how to reach every other host on that subnet, and goes to 192.168.0.254/24, the default router, to get to any other network. In this case, there is only one other network. The same rules apply on the 192.168.1.0/24 network.

How did the hosts know how to reach other local hosts, including the local interface of the default router? Before sending to a given host, they used the Address Resolution Protocol (ARP) to find the medium access control (MAC) address of that host, and simply used data link layer protocols to send to it.

How did they learn the address of the default router? There are many ways that could happen. In this case, assume that the hosts are preprogrammed with its address. It is quite possible to have backup default routers, transparent to the host, if the Virtual Router Redundancy Protocol (VRRP) is used, but that is a more advanced topic. Indeed, load can be shared among multiple default routers. No matter how many default routers, the topology is still stub on each side of the router.



Image:Routing-Stub-to-ROW.png|thumb|right|500px|Stub network defaulting to Rest of World In the second example, the focus is on the 192.168.0.0/24 network alone. Without loss of generality, it remains a stub network, still having to go to the default router for any address in a non-local network.

What is a non-local network, anyway? Using the CIDR notation, any network whose first 24 bits is not 192.168.0 is non-local. Using older terminology, the CIDR designation /24 is equivalent to the subnet mask 255.255.255.0.

In both cases, it is the default router's responsibility to know how to get to any and all other networks. With the first topology example, simple interface configuration tells the router that it is connected locally to 192.168.0.0/24 and 192.168.1.0. In the second topology, it could just be given a default route, 0.0.0.0/0 by convention, with a next-hop address of the access network that takes it to the rest of the world.

It could also have hundreds of interfaces to other stub and transit networks. 192.168.0.0/24, however, does not care.