Unused ESXi NICs alerting in vcenter

[soda9191]

I have a few different DL360 Gen 9s that are constantly sending flapping alerts on unused NICs. There are a total of 8 NICs on the server and only two of them are in use. I have done the following to try to remedy the situation

1. updated the firmware.

2. set the unused NICs to admin down on the ESX console.

3. changed the speed and duplex to full.

message in vcenter

Physical NIC vmnic7 linkstate is down.

event type id - esx.problem.net.vmnic.linkstate.down

vcenter version - 14070654

ESX version - 6.7 13981272

Is there anything else that can be done to stop the false alerts on the unused NICs?

[mc1903cae]

Did you install with the Custom HPE ESXi code or the ‘vanilla’ VMware ESXi code?

Have you seen the HPE Software Delivery Repository vibsdepot (aka HPE Online Depot)

I have installed ESXi on lots of HPE DL/ML’s and have not seen this type of error with unused NIC’s.

M

[KabirAli82]

Can you disable the NICs from the BIOS?

[a_p_]

Just to confirm, did you also update iLO to the latest firmware version?

There have been issues like this for at least Gen10 models with older iLO versions (in combination with the customized ESXi image).

André

[IRIX201110141]

Long time ago..... around ESX(i) 4.x i have to add new HP DLxxx servers to one of my larger customers. As a "Dell Guy" i found it not so difficult to rack&stack them and install ESX to them and attach them to the NetApp.  As i come back from the DC the Admin ask me if i place the BaseT blind adapters into the unused 1GbE LOMs.   WTF?!

If the customer doest place these adapter he also get the "not connected/down" warnings within his vCenter.  I never saw this when working with FSC, IBM/Lenowo or our PowerEdge. Also on recent DL380 Gen9/10 i havent seen this again. But today the server use 10/25G rather than the 1G from the past.

This comes into my mind after reading your posting.

Regards,
Joerg

PS: You can prove that you havent assign these unused VMNIC to an vSwitch right?

[soda9191]

The image was installed using the HP custom ISO. All of the firmware was updated including the ILO. There are only two assigned NICs to the distributed switch. I cannot disable the NICs per port in the BIOS. They would have to be disabled per card which would not be possible because the way we configured our HA.

[soda9191]

We are in the process of updating the ILOs to version 2.73. Will monitor and see if this corrects it. Our next step is to update the remaining firmware (We updated everything before with no luck).

I attached a screenshot of the NIC that is flapping. As mentioned there is no physical cable plugged into this 10GB port and it has been administratively been shutdown in the ESX console.