cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
556
Views
0
Helpful
2
Replies

C240 M4 VMWare host connected to 2 Nexus 35k in HSRP cannot receive routed packets and cannot ping default GW on the Nexus boxes

mdornfeld
Level 1
Level 1

I suspect I am missing something easy.

Two Nexus 35K switches in Active/Standby HSRP mode work perfectly for all all network routing except for 1 VMWare host.  The VMWare host has two 10GB vNICs connected to the pair of Nexus switches. Both vNICs are active. All routing is done on the Nexus boxes.

Our wireless network is able to connect to all hosts except to the hosts on this C240 VMWare 5.5 box. There are some other oddities such as a new VM added recently cannot access the Internet when default GW is the Nexus pair of switches. Other hosts have no problem.

Seems like the problem is a combination layer 2, layer 3 issue, but I can't really make that determination yet.

Nearly all other services work correctly.

Is this a VMWare vNIC configuration problem? 

Thanks in advance for any help offered.

2 Replies 2

Kirk J
Cisco Employee
Cisco Employee

Greetings.

Please confirm what kind of port channel (if any) you have configured on the nexus side connected to the eth ports going to the esxi host, and also confirm what kind of teaming configuration is set at the vswitch level.

Looking at mac-address table on both nexus switches confirm where the various guestVM macs are being learned.  Maybe disable one of the the VMnics in the vswitch (move it down to unused from active), and see if connectivity issues seem to stop. 

Check traceroute output from nexus switches to IP of guestVMs having issues, to make sure there's nothing odd with routing.  Can the guest VMs with DG issue successfully ping with all other devices on same vlan/subnet?

Thanks,

Kirk...

Port Channel is not configured on the Nexus side - the two ports are configured as access ports allowing a single VLAN.

The vmNICs are configured as Active-Active no teaming.

We did disable one of the vmNIC's with no change to behavior.

Traceroute from the active switch in the HSRP pair works properly.

VM's with DG issue can ping everything else on the network. 

I have been looking for a definitive setup in our situation.  I can't seem to sort out all the examples.

Another question I have is whether the vNIC's on the Cisco VIC 1225 are running true 802.11q and should behave like a switch trunk port.  A Cisco engineer said I have the ports set up correctly as trunks, but the Nexus ports are Access ports.  I explained that to him too.

Review Cisco Networking for a $25 gift card