Running VM kills Unraid network interface

[Sandstalker]

tl:dr Unraid network interface crashing when VM running

 

Morning All, hoping some of you smarter fellows can help me out.

I feel like this should be an easy fix but not having much Google luck.

 

UnRaid specifics:

- 6.9.2

- Plugins: Community Applications / Fix Common Problems / NVidia Driver / Unassigned Devices

- Docker: deepstack (AI for image analysis)

- Network config in Unraid is set with bonding off, bridging on, IPV4 only.

- Ethernet controller is not passed through

- Diagnostics attached

 

Hardware specifics:

- Motherboard = ASRock AB350M Pro4

- CPU = RYZEN 7 1700 8-Core

- Memory = Crucial Ballistix 3200 MHz DDR4 DRAM (32 GB)

- Drives = 2x 8TB, 1x 8TB Parity, 1x 240 GB SSD cache

- GPU = EVGA 970 GTX FTW

 

VM specifics:

- Win 10 Pro

- 4 cores

- 8GB Memory

- No hardware passed through

- XML attached

- VM config: Network Bridge is set to "br0", Network Model is "virtio-net".

 

Issue:

I recently swapped my cache SSD sata drive for an M.2. Upon starting up the server I found there was no connectivity between the server and router. I don't know that the drive swapped caused the issue but likely exposed it when the server was rebooted.

Server is set to dynamic; I have a static assignment setup on the router and found while the server was maintaining the IP (ifconfig), the subnet mask was pulling a /32 (255.255.255.255) instead of a /24 (255.255.255.0). After stopping docker & VM services and setting the IP to static on the server it cleared up the issue.

Fast forward some troubleshooting with the router, I believe I've narrowed down the cause of the issue to the VM running. The service can be running but as soon as the VM is started (even before Windows boots) the network interface on the server becomes unreachable. Not thinking it's the Windows OS on the VM but poked around a bit anyway. Found a 2nd adapter that was pulling the same IP as the server, disabled that as well as the primary for good measure, no luck.

The only way to get the interface back up and connected is to stop services and modify a setting in the network config. After applying everything is back to normal.

 

Any and all help is appreciated.

 

bitthurmzunraid-diagnostics-20220709-1130.zip VMBackup.7.9.2022.post.xml

[JorgeB]

14 hours ago, Sandstalker said:

No hardware passed through

NIC is being passed-through to the VM:

 

-device vfio-pci,host=0000:05:00.0,id=hostdev0,bus=pci.0,addr=0x8 \

 

05:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [10ec:8168] (rev 11)

 

 

 

 

 

[Sandstalker]

5 hours ago, JorgeB said:

NIC is being passed-through to the VM:

 

-device vfio-pci,host=0000:05:00.0,id=hostdev0,bus=pci.0,addr=0x8 \

 

05:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [10ec:8168] (rev 11)

 

 

 

 

 

You are correct, found this in the VM XML:

 

    <hostdev mode='subsystem' type='pci' managed='yes'>
      <driver name='vfio'/>
      <source>
        <address domain='0x0000' bus='0x05' slot='0x00' function='0x0'/>
      </source>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x08' function='0x0'/>
    </hostdev>

 

Though the VM editor UI didn't show any indication of it, nor did the "Tools > System Settings":

 

As I figured it was a simple fix, just took the right set of eyes and the right knowledge to see it.

 

Thanks @JorgeB!