Hello,
I'm facing some strange NIC problem on Windows 2008 R2 @ESXi 6.0
On W2008 VM there are 2 NIC's (both VMXNET3) + VMware tools installed. Both cards are assigned correctly in vSphere to VLANs. The problem is, that 2nd NIC card doesn't ping anything in it's subnet except the static IP address on this card. When I poweroff VM, and remove the 2nd NIC, add a new one (and assign it to proper vlan), assign the same static IP address - everything is OK... but only until VM restart. When VM is restarted - it will not ping any device in the same subnet.
I have 2nd W2008 R2 VM on the same ESXi host, and it has the same NIC configuration, and there it works well.
Just wonder, how to fix this problem permanently.
The problem is solved. Someone has changed physical link from server to the switch, and put it to the wrong port (non-trunk). ESXi has set "Route based on originating virtual port" load balancing algorithm, so the VM 2nd NIC was always bound to the "faulty" port.
If both VMs are on the same ESXi host, it can ping each other on the 2nd NIC but not if they are in different hosts?
Are you running NSX?
Maybe this is your case
Microsoft Updates replaces existing vmxnet3 vNIC on Windows Server 2008 R2 and Windows 7
Yes, these 2 VMs can ping each other, via their 2nd NIC.
Forgot to mention, I don't have any march updates installed. I've already tried removing old "ghost" NICs, but the problem still remain after 2nd VM restart.
Have you tried using E1000 NIC instead to see if it fixes the issue?
Yes, I've tried, it's the same on E1000.
Have you got any antivirus, host firewall, etc blocking the traffic? Is the 2nd NIC used for WLB or clustering?
The problem is solved. Someone has changed physical link from server to the switch, and put it to the wrong port (non-trunk). ESXi has set "Route based on originating virtual port" load balancing algorithm, so the VM 2nd NIC was always bound to the "faulty" port.