Nexus1000v Management stuck in lveth Port

[christoph.brunner]

Now this is a strange new thing I have...

After migrating some Storage, which worked fine. I lost connection to my Nexus1000v. Everything seems to work fine so the VM's are up and running. But I lost my SVS connection and worse I cannot manage my Nexus any longer.

On the console I've discovered that something with lveth (local virtual ethernet) is happening. With the command sh mac address-table vlan 229 (my Mgmt Vlan) I have the two Gateways out on the Port-Channel which is fine. Then I have four static entries with the two VSM mac addresses on lveth3/10 and lveth4/51 and the other VSM on lveth4/0 and lveth3/45...

I've set up a Layer 3 Interface on the physical switch where the ESX's are connected. There I can see the Nexus correctly in the IP ARP table...

Now I'm wondering if anybody had this problem as well?

[nenduri]

Is the VSM booting from local disk? or from a central storage?

You should see only one static entry for VSM's management MAC address and it should be a dynamic entry on others.

Looks like you have VSM HA setup.

If I understand, you are saying that you can reach/ping VSM by creating a L3 interface on upstream switch. Then it is something to do with network configuration on upstream switches..

Is the VSM behind N1K VEM? Can you send the output of the following commands on the host where your VSMs are hosted.

vemcmd show l2 <mgmt vlan>

vemcmd show bd <mgmt vlan>

vemcmd show port

vemcmd show trunk

thanks,

Naren