Unable to ping devices directly connected to Nexus 3548

[michaelgarcia2]

We have a Nexus 3548 switch that is the top of rack switch for our new flexpod. I'm trying to get our existing network to see the new subnets. I have setup a trunk port on the Nexus 3548 and on our Small business series Cisco switch. The Nexus can ping all of the new and old network including devices directly connected to it. When I try and ping devices on the Nexus (i.e. the NetApp, UCS, etc) I cannot. When I do a tracert form the pc on the old network to the nexus it shows me getting to the Nexus but the Nexus wont forward the traffic to the end points. Its stops once it hist the Nexus. The trunk interface going to our old switch has access tot hose vlans that the devices are on. Any ideas?

I have attached a Show Run if anyone wants to have a look. Thanks.

[Nuno Marques]

Hi Michael

im having the same exact problem, how did u solve it or did you find the root cause?

thanks mate.

[Nuno Marques]

Hi Michael,

can you advise if your problem was solved?

thx mate.

[Stefan Schonert]

Hi,

I'm having the same problem on a connected UCS (2.2.2c) via vPC on the Nexus 3548 (6.0.2.A3.1).

Is there any solution or further information on that problem?

Thanks!

[Nuno Marques]

Hi Stefan,

my case is still under analisys from TAC, they have reproduced the issue,

we also have removed VPC and we have problems  even connecting a standart L3 interface from the UCS and pinging the SVI on the directly connected N3K?!?

We have configured a static mac entrie on the svi as a workaround and it restored the connectivity...

real strange issue...let me know if you reach some conclusions.

Thanks!

[Stefan Schonert]

Hi nrmarques,

can you send me your SR# as reference?

I would contact our Cisco service rep directly so that I can point him into the right direction.

We've installed two exact same systems at the costumer (UCS+N3k) and just one is hitting this failure/bug.

Thanks!

[Nuno Marques]

Hi Stefan,

beware that thare are some scenarios where this happens that are related to a bug in the VIC cards, nevertheless this is not our case, in our case even a single link to the N3K has conectivity problems...

they have open a new internal bug for the issue but its still in analisys, it seems to have a WA that is to configure a vlan interface on the bond in OS side (regular trunk on the VPC N3k side) will test it next week.

what  UCS are u using? and are u using HSRP?

thx

regards

[Stefan Schonert]

Hi nmarques,

we also tested a single host connected to the N3ks and the error occured also.

UCS is dual 6248 and a brand new chassis witch 5x B200 M3 and Firmware 2.2.2c.

Regarding HSRP I have to get back to the costumer as we just installed the new systems over there.

regards

[Nuno Marques]

Hi stefan,

Cisco have open bug for the issue but its still in analisys:

CSCuq61825

Let me know if you had some progress on your problem.

thx

[Stefan Schonert]

Hi nrmarques,

it was another problem on the 3548 which resulted in the symptoms mentioned.

Our costumer had activated storm control on the VPC peer link with a level of 0.1 and all broadcast traffic was blocked by that.

So in my case it is a bug in the storm controll feature. Now reportet at:

CSCuq69820

"Broadcast traffic dropped when storm control configured with < 0.14%"

regards