by Matthew G. Naugle
Wiley Computer Publishing, John Wiley & Sons, Inc.
ISBN: 0471196568 Pub Date: 11/01/98
|
|
Illustrated TCP/IP
by Matthew G. Naugle Wiley Computer Publishing, John Wiley & Sons, Inc. ISBN: 0471196568 Pub Date: 11/01/98 |
| Previous | Table of Contents | Next |
From the previous discussion on choosing a subnet mask, a routing protocol known as RIP version 1, or RIP1, requires that a subnet mask be uniform across an entire network ID. An address assignment of 150.1.0.0 must contain one network mask. The fault here is the inability of RIP1 to supply a subnetmask entry in its routing updates to be consumed by other routers. Therefore, RIP1 is forced to make assumptions. It assumes that the mask is the same for the learned subnet of the same network ID as its configured ports. This means that if a subnet route is learned on a port that has the same network ID as the port, RIP will apply the assigned mask to that learned route as the port. If the learned subnet route has a different network ID than the port it learned the subnet route from, it assumes the learned subnet route is not subnetted and falls back to applying the natural mask for that class.
Here’s an example: A router has two ports. Port 1 is assigned an address of 150.1.1.1 with a subnet mask of 255.255.255.0. Port 2 has an address of 160.1.1.1 with a subnet mask of 255.255.255.0. If the router learns of a route 150.1.3.0, then it will apply the 24-bit subnet mask because it has the same network ID as its port. However, if the router learns a subnet route of 175.1.6.0, this network ID is not on either one of its ports and it will apply a natural subnet mask of 255.255.0.0 to that address before updating its table. That is for learned routes.
How about routing updates? When does a router apply the subnet mask to a route and then include it in the routing update? The same rule applies. Using the network numbers from the preceding example, when the router would like to broadcast its table, it will apply the subnet mask of 255.255.255.0 to the learned route of 150.1.3.0 when it sends its update out Port 1. However, it will send the address of 150.1.0.0 when sending the update out Port 2. Port 2 has a different network ID associated with that port and, therefore, the natural mask is applied before sending out the table.
This is why RIP1 supports only one subnet mask for network ID.
The next section gives more examples of address assignment and Variable-Length Subnet Masks (VLSM).
| Previous | Table of Contents | Next |
)