Is Tunnel interface on a router considered as a local interface or not?

[Jovica Todorovic]

Hi all,

We have DMVPN hub with about 100+ spokes and by chance we discovered that that there is a latency involved when pinging local tunnel interface on a spoke router.

i.e. on spoke rotuer:

XXXXX-RTR-01#sh ip int br
Interface IP-Address OK? Method Status Protocol
GigabitEthernet0/0 xxx.xxx.xxx.1 YES NVRAM up up
GigabitEthernet0/1 xxx.xxx.xxx.106 YES NVRAM up up
Tunnel12 10.200.2.48 YES manual up up

 XXXXX-RTR-01#traceroute 10.200.2.48
Type escape sequence to abort.
Tracing the route to 10.200.2.48
VRF info: (vrf in name/id, vrf out name/id)
1 10.200.2.1 118 msec 120 msec 118 msec
2 10.200.2.48 236 msec 238 msec *

XXXXX-RTR-01#ping 10.200.2.48
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.200.2.48, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 236/237/238 ms

It looks like ICMP paket is delivered from spoke to hub and returned back while pinging its own interface. With traceroute it looks the same.

Is this feature or bug?

Thanks

[Richard Burts · Accepted solution]

It is a feature. It has been a standard part of IOS for many years that for point to point interfaces like tunnels and some serial interfaces that when you ping the address of your interface that the ping is sent to the neighbor who forwards it back to you (and the same is true when your interface responds to the ping or traceroute that the response is sent to the neighbor who forwards it back to you).

HTH

Rick

[Jovica Todorovic]

Dear Rich,

many thanks.

Jovica

[Richard Burts]

Jovica

I am glad that my explanation was helpful. Thank you for using the rating system to mark this question as answered. This will help other readers in the forum to identify discussions that have helpful information.

HTH

Rick