We have had a few occurences of this problem, I would like to know if anyone has seen something similar or knows how to fix it;
In a single Distributed Switch, I have 8 uplink ports.
dvUplink1 and dvUplink2 are active uplinks for dvPortGroup_Console
dvUplink3 and dvUplink4 are active uplinks for dvPortGroup_Corporate
dvUplink5 and dvUplink6 are unassigned, planned for a future DMZ
dvUplink7 is an active uplink for dvPortGroup_DMZ
dvUplink8 is unassigned.
I added dvUplink8 to be an active uplink in dvPortGroup_DMZ.
Here's where the problem happened - we have 6 hosts in this dvSwitch, and on all of them the network card was unassigned from dvUplink7. This caused a loss of connectivity to all VMs in dvPortGroup_DMZ.
I found some of the vmnics assigned to dvUplink5, and on other hosts they were unassigned/unclaimed. I went through all the affected hosts and used Manage Physical Adapters to re-assign the vmnics to the right dvUplink to return connectivity.
I have had this problem when first deploying a new host - migrating from a Standard Switch and then re-assigning the vmnic to the dvUplink will result in other vmnics on other dvUplinks being unassigned and therefore a complete loss of connectivity requiring Restoring Standard Switch from the ESXi host console.
Has anyone else seen this problem or know the resolution?
