EIGRP default route propagation

cristi.grigore

Hi everyone,

I have the following scenario:

            R3
           / |
R1------R2   |
           \ |
            R4

R1 is connected to the outside world and I want to create a default route and advertise it to the other 3 routers through EIGRP. I did this setup on dynagen, so I had to create a loopback interface on R1 and use at as the default route to the outside.

The loopback has an ip address of 172.16.10.1/24 and it seems there is a problem with the fact that the ip default-network command is classful: whenever I enter the command ip default-network 172.16.10.0, the command changes into ip route 172.16.0.0 255.255.0.0 172.16.10.0 in the running config. I then have to use ip default-network 172.16.0.0 to mark the default candidate, instead of 172.16.10.0. The default candidate in the routing table is now 172.16.0.0 and it looks like this:

 *   172.16.0.0/16 is variably subnetted, 2 subnets, 2 masks
C       172.16.10.0/24 is directly connected, Loopback1
S*      172.16.0.0/16 [1/0] via 172.16.10.0
     10.0.0.0/24 is subnetted, 4 subnets
D       10.2.0.0 [90/30720] via 10.1.0.2, 00:43:19, FastEthernet1/0
D       10.3.0.0 [90/30720] via 10.1.0.2, 00:43:19, FastEthernet1/0
C       10.1.0.0 is directly connected, FastEthernet1/0
D       10.4.0.0 [90/33280] via 10.1.0.2, 00:43:19, FastEthernet1/0

Of course, under router eigrp i have the network 172.16.0.0 command in order to redistribute the default route.

The problem is: its not working, the default route is not advertised in the netw If I change the static route from ip route 172.16.0.0 255.255.0.0 172.16.10.0 to ip route 172.16.0.0 255.255.0.0 loopback 1, the default route will be advertised in the netw. Also, if I use a class C address on the loopback (192,168.10.1, for example) and set ip default-network 192.168.10.0, the default network of 192.168.10.0 is propagated.

Cisco says that:
"...(EIGRP) it redistributes a default route as a result of the ip route 0.0.0.0 interface command (but not as a result of the ip route 0.0.0.0 address..."

My opinion is that this is a bug in EIGRP, but Cisco put it in the documentation and it became a feature

What are your thoughts on this?

PS: here is the running config from R1 for the situation where it actually propagates the default route:

R1#sh run
Building configuration...

Current configuration : 868 bytes
!
! Last configuration change at 04:20:53 UTC Tue Sep 4 2007
!
version 12.4
service timestamps debug datetime msec
service timestamps log datetime msec
no service password-encryption
!
hostname R1
!
boot-start-marker
boot-end-marker
!
!
no aaa new-model
!
!
ip cef
!
!
!
interface Loopback1
 ip address 172.16.10.1 255.255.255.0
!
interface FastEthernet0/0
 no ip address
 shutdown
 duplex half
!
interface FastEthernet1/0
 ip address 10.1.0.1 255.255.255.0
 duplex half
!
router eigrp 1
 network 10.0.0.0
 network 172.16.0.0
 no auto-summary
!
ip default-network 172.16.0.0
ip route 172.16.0.0 255.255.0.0 Loopback1
!
no ip http server
!
!
control-plane
!
!
line con 0
 logging synchronous
 stopbits 1
line aux 0
 stopbits 1
line vty 0 4
 login
!
!
end

R1#

dtlokee

I don't think it's a bug in the IOS, just how the IOS behaves. Since the administrative distance of a static route pointing to a connected interface will be 0 and it is treated as connected, it can be originated with a "network" statement in router configuration mode. If the route is created pointing to another IP address, it will have an administrative distance greater than or equal to 1 so it would need to be redistributed into the routing process using a redistribute command.

mikej412

cristi.grigore wrote:

Cisco says that:
"...(EIGRP) it redistributes a default route as a result of the ip route 0.0.0.0 interface command (but not as a result of the ip route 0.0.0.0 address..."

My opinion is that this is a bug in EIGRP, but Cisco put it in the documentation and it became a feature

What are your thoughts on this?

Sometimes it is easier to change the documentation than to fix the program code.

Could have been poor programming, delivery deadlines, poor specifications, code size issues, etc...

At least they did document it

cristi.grigore

So what if R1 is connected via ethernet with the isp border router (one that has proxyARP disabled). This means I can't use a static route pointing to the exit interface. And this means I can't do this kind of default route redistribution via EIGRP.

mikej412

cristi.grigore wrote:

This means I can't use a static route pointing to the exit interface.

You still could, it just wouldn't get advertised as the default route..... so you'd probably do an ip summary-address on the R1 interface going to R2 for the 0.0.0.0 network -- that should get advertised to the rest of the AS.