X.25 can reroute around failures, which is not a characteristic of Frame Relay. One of the
reasons to use static routing is that some businesses bill for their X.25 packets. This often is the
case with airlines. Therefore, if you were to use a dynamic routing protocol such as IP to route
around failures, the broadcast packets would add up into billable packets. X.25 does not use
broadcast packets to locate resources.
cloud. If you do this, you will have to set up a bunch of static X.25 routes. You still want to
consider bandwidth arithmetic and some of the other factors, but you don't have to worry about
broadcast overhead. There are networks that have as many as 1000 sites in a star topology
using X.25.
Remember, the highest speed that an X.25 interface supports is 2000000 bits per second.
Usually X.25 circuits are very low-speed--typically 9.6 K or less. If you do the bandwidth
arithmetic, there could be approximately 200 9.6 K circuits to a central site with a 2,000,000-
bit X.25 interface. With lower-speed circuits, the number of remote sites can be even higher.
subinterfaces. Subinterfaces eliminate the need for split horizon issues in the network. This is
probably a good idea because it makes all the circuits look like point-to-point protocols, and
routing works correctly because point-to-point interfaces broadcast by default.
layer routing and full connectivity with a partial mesh.
by default.
to use routers as packet switches, but some companies are using legacy public X.25 networks,
and that can be very costly. If you have an alternate link in your network, you might consider
carrying the X.25 traffic on your regular data network instead of paying for someone else to
do it.