We had a failure on a core switch and following we got strange issue with our ESXi host that were running HP / Intel nic cards with driver ixgben. The state of the nic was Admin Down on ESXi side. On the switch it was up. FYI we were running IP Hash on the vswitch.
I will give you the workaround, first so you can fix your servers. But I will also explain the troubleshooting that I did. That bad thing is whats happens when you patch or do a new installation of ESXi you need to remember this. thats why this is a workaround and not a soloution. I hope HP/Intel will come with a working driver.
Workaround is to change the driver to ixgbe. to do that disable the ixgben with below command in the cli of the ESXi host. Then you need to reboot.
esxcli system module set --enabled=false --module=ixgben
Link about enable/disable driver in ESXi:
Command to to see the issue. The Admin down status is occurring when the port is back and up and running after a failure.
esxcli network nic list
vmnic4 0000:04:00.0 ixgben Up Up 10000 Full 38:ea:a7:27:13:2c 1500 Intel(R) 82599 10 Gigabit Dual Port Network Connection
vmnic5 0000:04:00.1 ixgben Down Up 10000 Full 38:ea:a7:27:13:2d 1500 Intel(R) 82599 10 Gigabit Dual Port Network Connection
I found a great article that explains my problem perfect, but not the soloution. The command was not in there. And we tried different version of the driver. No luck!
Te command to issue was this, but as I explained it does not work with 6.7
esxcli system module parameters set -m ixgben -a -p "VlanRemoveWorkaround=0,0"
You can also fix this manually, then when this happens you need to connect to every esxi host and issue below command. The X is reffering to the vmnic number.
esxcli network nic up -n vmnicX
I hope this help anyone out there!