I'm getting this message CONSTANTLY on my unRAID 6.9.2 server. From what I've researched so far, I feel like this should be taking my whole network down, but it's not. All of my Docker containers along with the WebGUI are working perfectly fine on the network, but my VMs are not (maybe related?).

 

Any help would be appreciated, thanks.

csssrv-diagnostics-20220121-0709.zip

Just as an update/bump:

- Booting into safe mode leaves me without network access or my array

- This problem persists before I start the array on normal mode

10 minutes ago, Vairfoley said:

- Booting into safe mode leaves me without network access or my array

 

That's very strange.  Can you grab diagnostics (you can do it from a command prompt) after booting into safe mode.  It will probably provide more information that will help us see what's actually going on.

1 hour ago, Squid said:

That's very strange.  Can you grab diagnostics (you can do it from a command prompt) after booting into safe mode.  It will probably provide more information that will help us see what's actually going on.

Sure! I really appreciate the help.

csssrv-diagnostics-20220129-1818.zip

Interesting.  The reason for why you have no network in safemode is because your normal boot method has

rcu_nocbs=0-11 pcie_acs_override=multifunction 

but without those entries on the safemode config, the wrong device gets isolated from the system

 

Unfortunately it doesn't help with the original problem.  And I've seen the messages off and on (and had them myself) but never to the same extent, and I'm not sure of what a resolution is.  Needs somebody with more networking experience than I have.

It could be a race condition, the bond interface is not ready yet while the bridge interface is.

Since you have a single interface eth0 assigned, you can disable bonding in the network configuration, it has no added value.

 

16 hours ago, Squid said:

Interesting.  The reason for why you have no network in safemode is because your normal boot method has

rcu_nocbs=0-11 pcie_acs_override=multifunction 

but without those entries on the safemode config, the wrong device gets isolated from the system

I actually don't really need that override patch anymore. Any way I can turn that off without losing networking?

 

16 hours ago, bonienl said:

It could be a race condition, the bond interface is not ready yet while the bridge interface is.

Since you have a single interface eth0 assigned, you can disable bonding in the network configuration, it has no added value.

I turned off bonding and now the error reads:

kernel: br0: received packet on eth0 with own address as source address

csssrv-diagnostics-20220130-1840.zip

Edited January 31, 20224 yr by Vairfoley
Formatted error message as code for better readability

9 hours ago, Vairfoley said:

I actually don't really need that override patch anymore. Any way I can turn that off without losing networking?

 

You would have to first deselect the what ever you've got isolated in Tools - System Devices, and then turn off ACS override in VM Settings.

 

After rebooting you would then isolate the applicable devices you need for passthrough in VM Settings and adjust the VMs as necessary

21 hours ago, Vairfoley said:

I turned off bonding and now the error reads:

 

Your Unraid server is correctly configured, it looks something external, perhaps a switch causing a loop.

You could try disconnecting (unused) cables to localize the source of the problem.

 

On 1/31/2022 at 4:34 PM, bonienl said:

 

Your Unraid server is correctly configured, it looks something external, perhaps a switch causing a loop.

You could try disconnecting (unused) cables to localize the source of the problem.

 

 

I had port mirroring enabled on the switch port my server was connected to... huge facepalm. No more messages in the log and my VMs seem to have network connection again! Thanks!

Edited February 8, 20224 yr by Vairfoley
Adding more information