VMware Cloud Community

ESXi gets disconnected when creating PortGroup on Distributed Swwith same VLAN but different options

Something somewhat catastrophic happened today.

I have 2 VMs VMS_01 and VMS_02, an Alpha VAX emulator, that need a second vnic on same VLANs that I have already configured on 2 portgroups (all default,  Promiscuous mode, MAC address changes and Forged transmit to Reject), one on VLAN 603 and the other on VLAN 604, named VLAN_603 and VLAN_604, respectively).

I created 2 portgroups with distinctive name (VLAN_603_VMS and VLAN_604_VMS) setting with MAC address Changes and Forged Transmit to Accept. So far, so good.

I added a NIC on each of my VMs, VMS_01 on portgroup VLAN_603_VMS, VMS_02 on VLAN_604_VMS.

Then I powered my VMs on and, suddenly, the network went down on the ESXi on which they were hosted, the iDRAC reports the Physical NICs really going down on the host.

My manager decided to reboot the host, although I said it was just going to transport the problem to another host... which it did.

With an iDRAC session, I killed the 2 VMs with vim-cmd, restarted the agents, and the network came back instantly.

Because all the VMs on the host were running perfectly, although without any network, I take it's normal HA didn't kick in.

However, I don't understand why creating these 2 portgroups disabled my physical NICs on my hosts. Usually, I create such "aliases" on a separate Standard Switch, but I'm not sure what these different options on the PortGroups may have done. Or could it be something in the Physical Switches (Cicsco Nexus) that may have gone wrong?

Any idea? Obviously, I'm waiting to understand what happened before I try again, and that's a blocker for the Alpha Migration project...

0 Kudos
0 Replies