Network connectivity restored on virtual switch - but never went down?

[NateDreier]

Hello guys,

Let me briefly apologize in advance as I am not a vmware engineer and don't understand the intricacies of vmware.

I stumbled across this issue trying to troubleshoot why our Windows deployment server was crawling during the initial TFTP download of files. As I was trying to rule things out I came across this within ESXI itself.

On our ESXI host I have been noticing events stating -

Network connectivity restored on virtual switch "xxxx", protgroups: "xxxx", "xxxxxxx". Physical NIC vmnic0 is up.

However, there are no logs stating that the physical NIC went down.

This happens fairly frequently, at least every hour. It also seems to happen when VM's are rebooted as well as PXE'ing.

When I look in vobd.log I see -

[netCorrelator] 1234567us: [vob.net.pg.uplink.transition.up] Uplink:vmnic0 is up. Affected portgroup: XXXX 1 uplinks up

[netCorrelator] 1234566us: [esx.clear.net.connectivity.restored] Network connectivity restored on virtual switch "xxxx",portgroups: "xxxx" "xxxxxxx". Physical NIC vmnic0 is up.

This is happening roughly every 10~ minutes.

Again, I see no correlating "Physical NIC went down" logs.

EDIT: a few details I forgot to enter -

ESXi 6.7 update 1

Driver bnxtnet

(I will add as I think of things or things are asked)

Thanks in advanced. Hopefully this is a no duh no brainer issue.

Nate

[daphnissov · Accepted solution]

When I've seen this in the past, it has been due to network driver issues. First thing I'd check (in addition to patching your host) is if a newer version of that driver is available and supported for your NIC. You may also choose to open an SR with VMware and supply logs to see if this is a known issue otherwise.

[NateDreier]

EDIT: Final update. There was this firmware update that resolved EVERYTHING. PXE'ing (TFTP issues) went from 2 hours to 30 seconds etc. Changing the 'correct' answer over to Daphnissov.

Broadcom NetXtreme-E Network Device Firmware 21.4.1 | Driver Details | Dell US

Hey thanks for the reply!

Updated that to no avail regarding the log - however and I should have looked here first, I am running 6.7 U1, in the fixes for U2 they state:

PR 2256771: You might see an unnecessary event message for network connectivity restored on virtual switch when powering on or off virtual machines

In the Events tab of the vSphere Client or the vSphere Web Client, when you power on or off virtual machines, you might see events similar to Network connectivity restored on virtual switch XXX, portgroups: XXX. Physical NIC vmnicX is up. When checking uplink status, the system report might include port groups that are not affected by the virtual machine powering on or off operation, which triggers the event. However, the event does not indicate a problem or prompt an action.

https://docs.vmware.com/en/VMware-vSphere/6.7/rn/vsphere-esxi-67u2-release-notes.html

So updating to U2 should resolve this issue!