Redistribution

Redistribution
We briefly discussed redistribution in Chapter 8, “Border Gateway Protocol.” To refresh you,
redistribution is the process of allowing routing information known in one routing protocol to
be shared with another routing protocol. It should be noted here that routing protocols are the
only items that can be redistributed into. You can redistribute such items as connected interfaces,
static routes, and default routes into a routing protocol.
There are two types of redistribution available: one-way redistribution and mutual redistribution.
One-way redistribution occurs when routing information contained in a routing protocol
is shared with another routing protocol, but the other routing protocol doesn’t share its
routing information with the sharing protocol. Mutual redistribution is the process of two routing
protocols sharing their routing information with each other.
Before we get too far into redistribution, we should revisit some of the topics discussed earlier
in this study guide. If you can remember back to the first chapter, we discussed how routes were
selected for the routing table. Administrative distance is the first deciding factor in selecting
which route to place in the routing table. The route with the lowest administrative distance is
the route that will be selected. If you do not remember the administrative distance values, this
would be a good time to go back and review. There is a chance that more than one route to the
same destination will have the same administrative distance. When this occurs, the route with
Redistribution 337
the lowest metric is the route that will be selected; if load balancing is in use multiple routes may
be selected. Each routing protocol calculates the metric in its own way.
Another topic that’s important to review is the difference between classful and classless routing
protocols. Classful routing protocols do not advertise the subnet mask of a route. So either
received routes are handled as classful routes, meaning that they are assumed at the classful
boundaries, or they take the subnet mask of the subnet they were received on, if they share the
same major network number. Classless routing protocols send the subnet mask with the route.
This allows for better use of addressing. One of the problems you will run into with redistribution
occurs when you need to redistribute from a classless routing protocol to a classful routing
protocol. We will cover this topic in more detail later in this chapter.
Now would be a good time to go back and briefly revisit all of the IGPs that we have discussed
thus far:
RIP If you remember from our earlier discussion, there are actually two versions of RIP: RIPv1
and RIPv2. RIPv1 is a classful routing protocol, whereas RIPv2 is a classless routing protocol.
Both versions of RIP use a metric known as hop count. Hop count is the number of routers that
a packet must pass through to reach its destination.
IGRP and EIGRP IGRP and EIGRP are Cisco proprietary routing protocols. Both use a composite
metric made up of the following:
 Bandwidth
 Delay
 Reliability
 Load
IGRP is a classful routing protocol, whereas EIGRP is a classless routing protocol. Remember:
If IGRP and EIGRP are running on the same router and both of them are using the same autonomous
system number, they will automatically redistribute with each other. If they do not have
the same autonomous system number, you will need to manually redistribute between them.
These are the only IGPs we will discuss in this study guide that behave in this manner.
Although redistribution occurs automatically from IGRP to EIGRP when
identical AS numbers are used, the resulting route entries in EIGRP do not
have the same administrative distance as EIGRP routes originated within
the EIGRP AS. While native routes carry an administrative distance of 90,
routes redistributed into EIGRP, even those from the same AS in IGRP, carry
an administrative distance of 170, far less trustworthy or desirable. The converse
is not true, because IGRP has no default mechanism for favoring
native routes over redistributed routes to the same destination network. As
you will see in this chapter, however, you can influence the use of redistributed
routes by how you set their metrics.
OSPF and IS-IS OSPF and IS-IS are both classless routing protocols. Both of these protocols
also use the Dijkstra algorithm to calculate the cost of a link.
338 Chapter 10  Route Optimization