Static route using interface as target - why working with FastEthernet?

[Lukas Hubschmid]

Hello all,

I have build up a routing scenario (in Packet Tracer) and was surprised that it's working ...

My thoughts:

When adding a static route, it is allowed to use a Interface as target instead of next hop IP address.

In case of a serial connection between two routers, this is absolutly clear, as this is always a point-to-point connection (or am I wrong?).

Now: I was wondering WHY this is working also with FastEthernet-Interfaces ...

My scenario:

I have used 3 private networks (192.168.0.0/24, 192.168.1.0/24, 192.168.2.0/24), which are connected using 2811-routers through the network 10.0.0.0/8 (see also the attached image).

On the routers, I have added two static routes to the other private networks (e.g. to route from 192.168.0.0/24 to 192.168.1.0/24 and 192.168.2.0/24) using the interface fa0/1 for BOTH routes!!

I have NOT configured any dynamic routing protocol (RIP, OSPF, EIGRP).

My problem:

Routing is possible without any problems!!??

Why the heck is this working??

The routing tables shows only the two directly connected networks and the two static routes which are using interface fa0/1 as target.

No entries of dynamic routing protocol!

How does the router knows the correct MAC-address of next hop router, wihtout knowing the IP address?

Hope you can answer!

BR,

lousek

[Edison Ortiz · Accepted solution]

What gave you the impression it wouldn't work? While it's not recommended, fastethernet as the next hop is a working command:

From:

http://www.cisco.com/en/US/docs/ios-xml/ios/iproute_pi/command/iri-cr-a1.html#GUID-24399FED-0F6A-479C-B88A-D7FB74257058

The practical implication of configuring the

 ip route 0.0.0.0 0.0.0.0 ethernet1/2

command is that the router will consider all of the destinations that  the router does not know how to reach through some other route as  directly connected to Ethernet interface 1/2. So the router will send an  ARP request for each host for which it receives packets on this network  segment. This configuration can cause high processor utilization and a  large ARP cache (along with memory allocation failures). Configuring a  default route or other static route that directs the router to forward  packets for a large range of destinations to a connected broadcast  network segment can cause your router to reload.

Specifying a numerical next hop that is on a directly connected  interface will prevent the router from using proxy ARP. However, if the  interface with the next hop goes down and the numerical next hop can be  reached through a recursive route, you may specify both the next hop and  interface

(for example, ip route 0.0.0.0 0.0.0.0 ethernet 1/2 10.1.2.3)

with a static route to prevent routes from passing through an  unintended interface.

[Jose Jara · Accepted solution]

Hi,

It works due to Proxy ARP. The router will send an ARP Request for every destination, because it thinks that the remote network is directly connected as it has an interface instead of next hop IP address, then the router who knows to reach the remote network will reply with his own mac address to the ARP Request.

Best Regards,

Jose.

[AshmarinD_2 · Accepted solution]

It works because two things happens:

First, the router who has static route pointing to the interface sends ARP request for every packet with unknown destination.

Second, the

ip proxy-arp

command is enabled by default in routers. When router gets the arp request with destination for network that it has in its routing table it sends reply with his mac address in ARP packet for this destination.

You can check this with

 "show arp" and "show run all | s interface FastEthernet0/1" 

commands.

[Rahul Kukreja]

Check this Example -

http://ciscoexpert.wordpress.com/2008/06/28/proxy-arp/

and look for

sh ip arp

especially

[Lukas Hubschmid]

Aaaaah, proxy ARP was the missing keyword

Thanks very much to all of you!!

I think all your answers are "correct answers"

Again, thank you!