Mojave VM Ethernet failures

[bland328]

Aside from an Ethernet issue I haven't solved, I've got what seems to be a rock-solid Mojave VM running under Unraid 6.6.6.

 

My Ethernet issue is simply that often (but not always) after I boot the VM, I don't have Ethernet access. In this case, System Preferences > Network says that my Ethernet cable is unplugged.

 

If I then issue

$ sudo ifconfig en0 down && sudo ifconfig en0 up

from the Terminal, Ethernet starts working, and I've never seen it then fail again until I reboot the VM.

 

My network configuration looks like this (MAC address censored):

...
    <type arch='x86_64' machine='pc-q35-3.0'>hvm</type>
...
    <interface type='bridge'>
      <mac address='99:88:77:66:55:44'/>
      <source bridge='br0'/>
      <model type='e1000-82545em'/>
      <address type='pci' domain='0x0000' bus='0x01' slot='0x00' function='0x0'/>
    </interface>
...

Any thoughts on how I might fix this? Or, at least, gather more data?

Edited December 19, 2018 by bland328
Added my 'pc-q35-3.0' machine type

[bland328]

In this 2017 post:

@llonca13 says:

Quote

...if anyone has a problem with Ethernet showing disconnected...edit your xml under ethernet device - address type - put the slot number higher (for example slot=0x05)...EDIT: just seen that gridrunner already pointed out this in another thread.serves me right for not reading everything first  

That sounds promising! Though I haven't been able to find the original @SpaceInvaderOne (formerly @gridrunner) post referenced, I have attempted to change the slot number in the XML as follows:

...
      <address type='pci' domain='0x0000' bus='0x01' slot='0x05' function='0x0'/>
...

However, when I make just the single-digit change and then click Update in the XML View, I immediately get this error pop-up from Unraid:

 

VM creation error
XML error: Invalid PCI address 0000:01:05.0. slot must be <= 0

 

Any thoughts on what I'm doing wrong? Do I need to somehow update the definition of PCI bus 1, or otherwise "create" slot 5?

 

Or am I just all confused because @llonca13's original post above is only about passing through a physical NIC, and wouldn't apply to a virtual bridge? 

Edited December 19, 2018 by bland328

[1812]

works fine on mine. don't worry about the model shown or alias/etc..., I was testing something out the vmxnet3... also this is from a running vm. Whatyou should note is the address.

 

 <interface type='bridge'>
      <mac address='52:52:00:4h:1a:92'/>
      <source bridge='br0'/>
      <target dev='vnet4'/>
      <model type='vmxnet3'/>
      <alias name='net1'/>
      <address type='pci' domain='0x0000' bus='0x02' slot='0x00' function='0x1'/>
    </interface>

 

[alfredo_2020]

I’m having the same sporadic problem hasn’t happened that often but not sure if I’m just missing the model number cause I keep playing with trying to get more cores added. I use the e1000 device model value not the vmxnet3. I’m gonna try that and see if it helps.

 

also how many cores/threads can you pass to this OS X VM IT LOOKS LIKE ITS only boring if I have 4c/8t max selected.

[1812]

4 hours ago, alfredo_2020 said:

I’m having the same sporadic problem hasn’t happened that often but not sure if I’m just missing the model number cause I keep playing with trying to get more cores added. I use the e1000 device model value not the vmxnet3. I’m gonna try that and see if it helps.

 

also how many cores/threads can you pass to this OS X VM IT LOOKS LIKE ITS only boring if I have 4c/8t max selected.

I’d you remove the topology definition from the xml you can do any core count up to 64

[bland328]

For anyone else chasing this broken-Ethernet-at-boot issue, I seem to have fixed it on my system by changing the bridge PCI bus to '0x00' and slot to '0x05' thusly:

...
<interface type='bridge'>
...
      <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/>
...

Since making the change, over a couple dozen boots, I haven't again seen macOS report the (virtual) Ethernet cable as being unplugged.

[ghost82]

Thank you, I was playing around this issue recently and I found this post/solution; in my case increasing slot from 0x02 to 0x08 solved the issue (I have 2 bridges at bus 0x08).

Moreover, I found that the issue is back again if multiple networks are flagged as "built-in", so ensure that only en0 is flagged as built-in.

Edited May 4, 2020 by ghost82

[xxbigfootxx]

On 4/29/2020 at 11:07 PM, ghost82 said:

Moreover, I found that the issue is back again if multiple networks are flagged as "built-in", so ensure that only en0 is flagged as built-in.

How do you specify which networks are flagged?

 

My issue is that i'm unable to contact the apple recovery server. I've tried like a hundred times now, changing small details here and there but same issue. I noted that i wasn't able to resolve DNS and not sure how to set that in mac OS terminal or from the VM tab in unraid. 

[ghost82]

3 hours ago, xxbigfootxx said:

My issue is that i'm unable to contact the apple recovery server.

Ensure that you have vmxnet3 as virtual network, and change to e1000 only after downloading from server.

In opencore you add the built-in property in the "DeviceProperties --> add" section, by specifying the address of the ethernet and the buil-in property set to true.

I don't remember in Clover, I think that it can be similar, maybe also the FixLan does the same.

[xxbigfootxx]

12 minutes ago, ghost82 said:

Ensure that you have vmxnet3 as virtual network, and change to e1000 only after downloading from server.

This is my current XML file: 

    <interface type='bridge'>
      <mac address='52:54:00:a4:47:7b'/>
      <source bridge='br0'/>
      <model type='vmxnet3'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/>
    </interface>

 

Even using the default settings from @SpaceInvaderOne VM innstall guide it seems to result the 'recovery server error'.