Route Redistribution Between BGP and IGPs

As you are aware, a router running the BGP routing protocol handles its own routing table separately from the IP routing table. An AS uses three primary methods to inject route information into BGP: using the network command, redistributing static routes to null 0, and redistributing dynamic routes into BGP.

Using the network Command

First, the network command is used to manually configure BGP to advertise a network that already exists in the IP routing table. The list of network command entries must contain all the networks in the AS that you want to advertise. For example:

 RouterA(config) router bgp 65410 RouterA(config-router)# network 192.168.10.0 RouterA(config-router)# network 10.10.0.0 

If summary routes need to be advertised into BGP, then the redistribution of static routes can be used to bypass the requirement that the route exist in the IP routing table. One technique is to use the network command. The network router configuration command advertises a particular IGP network and generates a corresponding entry in the router's BGP routing table.

Know the basic definition of redistribution. The redistribution process happens when a router is running multiple routing protocols and it advertises the routing information that is received between the protocols. The process enables routing information learned from one protocol to be "disseminated" in the update messages of another routing protocol.

Routes that are advertised by the use of the network command are shown with the origin code IGP (I) in the BGP routing table. The preferred method for advertising a summary route, however, is via the aggregate-address router configuration command. The aggregate-address command generates a single route entry for a range of network addresses if at least one specific route entry that is contained within the aggregate address range exists in the BGP routing table.

As discussed in Chapter 2, "Routing Principles," classless interdomain routing (CIDR) is a mechanism developed to help alleviate the problem of exhaustion of IP addresses and the growth of routing tables. The idea behind CIDR is that blocks of multiple addresses can be combined, or aggregated, to create a larger classless set of IP addresses. These multiple addresses can then be summarized in routing tables, resulting in fewer route advertisements. Earlier versions of BGP did not support CIDR; BGP-4 does, however, and BGP-4 support includes the following:

• The BGP update message carries both the prefix and the prefix length (len). Previous versions of BGP included only the prefix, and the length was assumed from the address class.

• Addresses can be aggregated (summarized) when advertised by a BGP router.

• The AS- path attribute can include a combined unordered list of all ASs through which all the aggregated routes have passed. This combined unordered list should always be measured to ensure that the route is loop-free and stable.

Two BGP attributes are related to aggregate addressing. The well-known discretionary Atomic-aggregate attribute tells the peer AS that the originating router has summarized the routes. The optional transitive Aggegator attribute designates the BGP router ID and AS number of the router that carried out the route aggregation.

By default, the aggregate route is advertised as coming from the AS that did the aggregation and has the Atomic-aggregate attribute set to show that information may be missing. The Autonomous System numbers from the non-aggregated routes are not included in the list. The router can be configured to include the unordered list of all autonomous systems contained in all paths that are being summarized. Aggregate addresses are not used in the Internet as much as they could or should be because most multihomed ASs want to make certain that their routes are advertised, without being aggregated into a summarized route.

Redistributing Static Routes into BGP Using Null 0

Another recommended method for advertising an IGP network into BGP is to redistribute static routes into BGP by using the null 0 interface. Creating a static route by using the null 0 interface advertises a single aggregate address for IGP networks instead of multiple single IGP addresses for each network. This is actually a way of tricking the process into thinking that a route actually exists for the aggregate. (If you use the network command for a network that already exists in the IP routing table, the static route to null 0 is altogether unnecessary.) Although routing to null 0 literally means "discard the packet to this network," pointing a static route to the null 0 virtual interface is acceptable. The BGP router already has specific routes to all networks inside the local AS and ignores the static route in favor of a more specific path.

You create a static route with a null 0 interface by using the ip route global configuration command and the redistribute static router configuration command. For example:

 RouterA(config)# router bgp 65410 RouterA(config-router)# redistribute static RouterA(config-router)# exit RouterA(config)# ip route 0.0.0.0 0.0.0.0 null0 

Redistributing Dynamic IGP Routes into BGP

Although not really recommended, a third option is to redistribute dynamic routing IGP routes into BGP. The redistribution of dynamic IGP routes into BGP should rarely be configured because subsequent changes in IGP routing information may not be indicated in the BGP updates. A BGP update is generated every time a change is detected in the IGP route topology, such as when a link becomes unavailable. This increase in BGP routing updates may lead to instability in the BGP routing tables within an AS. Only local routes should be distributed. To avoid routing loops , routes learned from other ASs must not be re-sent from the IGP. Using a redistribute command into BGP results in the (?) incomplete Origin attribute entry for the BGP route to show up in the show ip bgp output. A sample of show ip bgp output is shown in Listing 9.5

Listing 9.5 A Sample show ip bgp Output

 RouterA show ip bgp BGP table version is 583644, local router ID is 172.16.157.32 Status codes: s suppressed, d damped, h history, * valid, > best, i  internal Origin codes: I  IGP, e EGP, ?  incomplete     Network        Next Hop           Metric      LocPref   Weight  Path *>  10.1.0.0       172.20.36.2             0          100        0  480  i *>  10.2.0.0       172.24.20.1             0          100        0  140  i *   10.3.0.0       172.23.1.10             0          100        0  340  i *>                 192.168.2.1             0          100        0  200  i *                  192.168.3.1             0          100        0  100  i RouterA(config-router)# exit RouterA# 

Looking at Listing 9.5, RouterA sends traffic destined for network 10.3.0.0 to the next-hop router 192.168.2.1 (notice the > symbol) because it determines that the path through AS 200 is the optimal path. In a BGP routing table, the first column displays the status of each stored route based on the codes in the legend above the table. Even though BGP can store several routes to the same destination, a BGP router has only one best, or preferred, route. The optimal path for each route is specified by a > symbol in the first status column. Table 9.2 describes each field found in the BGP routing table.

Table 9.2. The BGP Routing Table As Seen Via the show ip bgp Command

BGP Routing Table Field	Description
BGP table version	An internal version number that is incremented each time a routing change occurs.
Local Router ID	An internal identification number for the router; derived from the highest IP address configured for the router, often a loopback address.
Status Codes	Identifies the status of each route in the routing table: Suppressed ”Inactive route. Valid ”Active route. Best ”Best route to that network. Internal ”Route learned from the IBGP peer. Damped ”Route is currently reachable and is used, but has an instability metric from previous bad behavior. History ”Route has been withdrawn and is currently unreachable.
Origin Codes	Identifies the source of each route in the routing table. IGP ”Source of route originated with IGP and was advertised by means of the network command EGP ”Source of the route originated with EGP ? ”Source of the route is unknown, most likely distributed into BGP by means of IGP
Network	IP address of the destination network.
Next Hop	IP address of the next router along the path to the destination network.
Metric	The inter-autonomous system metric value.
LocPref	The configured local preference attribute for this route entry.
Weight	The configured weight attribute for this route entry.
Path	Listing of AS(s) traversed on the way to the destination network.