Greetings!
Hoping someone can help me out with this, I've spent hours on it....
This is for my homelab so some hardware is unsupported.
I have two OptiPlex 7080s with mostly identical hardware. Same memory, CPU, NICs, just some different local drives. Call them esx1 and esx2. I have combed through the BIOS on both, which is identical and up-to-date as well, and everything is set the same. I have a problem with esx2 where one of the ports on the X520-DA2 card just drops. It stops passing all traffic though the managed switch it's connected to shows it has link. ESXi says the link is down. A reboot of the host will bring it back up, but then it drops again within a couple of days at most.
I have reviewed this vmware article (can't link it) as it sounds similar: "Physical network connectivity lost on servers with activated ASPM (2076374)"
I have verified that ASPM is disabled on both hosts as per the link above. I have swapped the identical SFP+ DACs. I have swapped switch ports. I have swapped the x520 out with another. The problem is always the same port, which leads me to believe it's an issue with ESXi. I did have to edit the firmware with ethtool to allow non-intel DACs, but I did that to all 6 of these cards I have and all the rest are fine. esx1 never has a problem it's always with this esx2. Both are freshly installed and running VMware ESXi, 7.0.3, 20842708. Running "esxcli network nic get -n vmnic0" returns identical results on both hosts.
The only difference I can find is that on esx1, it says SR-IOV is "not capable" for both X520s under PCI Devices. On esx2, it shows SR-IOV as disabled, but configurable. I have no idea if this is relevant, and I have no idea why on one system it's not capable but on the other it's just disabled.
Anyone have any clue before I flatten this esx2 host and start over? I bought both of these and set them up identically, so I'm having a real hard time discovering why they appear to have some differences.
Thanks in advance.
