Loopback connectivity

[Muzaffar Iyad Boukai]

Dear

I have core switch with vlan 1 configured , and access switch that is connected to core switch .

core config

int vl1 1

ip add 10.31.1.1 255.255.255.0

int loop 100

ip add  10.70.1.1 255.255.255.255

sho ip route 10.70.1.1

Core1#sho ip route 10.70.1.0
% Subnet not in table
Core1#
Core1#sho ip route 10.70.1.1
Routing entry for 10.70.1.1/32
  Known via "connected", distance 0, metric 0 (connected, via interface)
  Routing Descriptor Blocks:
  * directly connected, via Loopback100
      Route metric is 0, traffic share count is 1

Core1#

--------------------------------------------------------------------------

access switch

int vl 1

ip add 10.31.1.200 255.255.255.0

!

in loo 100

ip add 10.70.1.200 255.255.255.255

CAB1-sw1#sho ip route
CAB1-sw1#sho ip route
                    ^
% Invalid input detected at '^' marker.

CAB1-sw1#

now what is confusing :

Core1#
Core1#ping 10.70.1.200

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.70.1.200, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)
Core1#

==================================================

CAB1-sw1#ping 10.70.1.1

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.70.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/8 ms
HO-CAB1-sw1#

can any one explain this please

regards

[Jon Marshall]

The short answer is your access switch is L2 so it simply sends all traffic to the it's default gateway which is the core switch (see below though) and that is why it can ping the loopback.

The other way round doesn;t work because the core switch has no route for the access switch loopback.

However there are some confusing things in your post -

1) the vlan interfaces are in different IP subnets - is that a typo ?

If it isn't then it works probably because you have no "ip default-gateway x.x.x.x" command on your access switch and the core switch has proxy arp enabled

2) the loopback 100 on your access switch - is it even up ie. what does a "sh ip int br" show on your access switch ?

Jon

[Muzaffar Iyad Boukai]

hi

ok , when the switch send the traffic to default gateway , although no gateway is configured on the access switch , the ping is echo request and echo reply , so when I do a ping 10.70.1.1               from the access switch and the ping Is successful then the request and the reply are working .

Then why when I do a ping from the core its not working?? keep in mind that the ping reply is working !!!

- its a type mistake both vlan 1 in the same subnet

- its a loopback interface it should be always up

CAB1-sw1#sho ip inter b
Interface              IP-Address      OK? Method Status                Protocol
Vlan1                  10.31.1.101     YES NVRAM  up                    up     
Loopback100            10.70.1.101     YES manual up                    up

[Jon Marshall]

The reason it works is the access switch has no default gateway. It is L2 so it is not routing. So it acts like a host.

Because it has no default gateway when it requests the mac address for 10.70.1.101 the core switch responds because it's vlan interface is running proxy arp (presumably).

The access switch then sends it's ping packet to the core switch with the destination mac address of the core switch vlan interface.

So the packet gets to the core switch.

The other way round though is different. The core switch is routing so when you ping the loopback of the access switch it looks in it's routing table, doesn't find an entry and therefore your ping fails.

There are a few things you can experiment with so you can see how it all works but is this a live production environment or a test setup ?

Jon

[Muzaffar Iyad Boukai]

thank you jon for your reply

I will quote

"The access switch then sends it's ping packet to the core switch with the destination mac address of the core switch vlan interface."

this is the echo request . it reaches the core switch and then the loopback , the core will perform the echo reply (to perform the ping)  

how is the echo reply is happing from the core as we agree that core is doing routing and there are no route for the 10.70.1.0 in the routing table.

I m not talking now about initiate the ping from the core, im asking about the icmp echo reply that is happing from the core to reply the icmp request from the access switch .  

hope I made my question clear :)

regards

Muzafar

[Jon Marshall]

Muzafar

how is the echo reply is happing from the core as we agree that core is doing routing and there are no route for the 10.70.1.0 in the routing table.

The core does have a route for the loopback, look at your original post where you do a "sh ip route 10.71.1.1".

Jon

[Muzaffar Iyad Boukai]

yesss and that's why I m confused

CAB1-sw1#ping 10.70.1.1

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.70.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/8 ms
CAB1-sw1#

this means icmp echo request and echo reply is successful working !!!!

[Jon Marshall · Accepted solution]

Muzafar

Which bit are you confused by ?

The core switch receives the echo request and it has the loopback interface configured.

So the echo request is successful.

The source IP of that echo request will be the vlan interface IP on the access switch. The core switch knows how to get to that as well.

So the echo reply is successful as well.

I'm not sure what part of it you are not understanding.

Jon

[Muzaffar Iyad Boukai]

hmmm , ya I totally forgot about the sours interface :(

thank you jon that was a dummy question from me