Switching

Hello , guys , as the tittle says:

Im having some problems with the configuration posted bellow:

Im able to ping :

From 172.22.0.254 (mx960) to 172.22.0.253 (me3400) on vlan2 and

From 172.23.0.254 (mx960) to 172.23.0.253 (me3400) on vlan3

Without any packet loss.

then , from:

From 172.22.0.254 (mx960) to 172.22.0.1, host on vlan2 and

From 172.23.0.254 (mx960) to 172.23.0.1, host on vlan3

I have an %22 packet loss average.

Very strange:

I think it could be related to a mismatch encapsulation issue.

btw , my 3400 does not support the "switchport trunk encapsulation dot1q" directive.

Any ideas?

Ok , here is the scenario:

Also there are some messages at me3400 console:

.May 28 10:03:39.917: %SW_MATM-4-MACFLAP_NOTIF: Host 7819.f7bd.e7f0 in vlan 2 is flapping between port Gi0/8 and port Po2
.May 28 10:03:55.092: %SW_MATM-4-MACFLAP_NOTIF: Host 7819.f7bd.e7f0 in vlan 2 is flapping between port Gi0/9 and port Po2

After reading a while , Im thinking about a layer2 loop. but .... How is that possible ?

Between my MX960 and the Me3400 is only one possible path.

7819.f7bd.e7f0 is the irb mac at MX960 ... so could the loop be on the interfaces within the link agregation?

Can you correlate the packet loss with the logs seen on Cisco console?

Is the packet loss constant?

Can you disable one physical link at the time and see if the problem is still there?

=====

If this worked for you please flag my post as an "Accepted Solution" so others can benefit. A kudo would be cool if you think I earned it.

have you tried testing the ping form Cisco to MX and then from host to MX?

Also is it possible to create trunk interface on MX going to CISCO?

You want to monitor the interface to see what kind of packet losses are occurring?

It could be that your LAG is not set up properly between the two devices.  Can you post the configurations of the member interfaces of both 'ae6' on the MX960 and 'Po2' on the ME3400?