Hi,

So, first up, apologies if wrong forum etc. etc. I don't know if it's a networking issue or a VM one, so I figured I had to start somewhere...

This is (at least as far as I'm concerned) a really odd one...

I have a VM on KVM on my Unraid 6.5.3 system which is running Ubuntu and I use as my Mail Server. (I intend to move it over to Docker, but that's a project for another day).

Anyway, I am using a bridged network and I have a single subnet for my entire network.

I have a variety of other machines (physical) including a couple of Windows 10, an Ubuntu-based router/firewall and some iOS devices.

So, if I go to either my host machine, or my Ubuntu router, I can ping the VM in question (and vice-versa). From the Windows machines, I cannot (times out). Inbound e-mail gets delivered to the email server (via the firewall) and outbound e-mail gets sent. I can check my e-mail from my iPhone, but not from my windows desktop.

On the Host, I keep getting the error "Tower kernel: br0: received packet on eth0 with own address as source address addr:9c:8e:99:0b:e2:0a, vlan:0)" in syslog.

This is driving me completely insane! Can anyone suggest what the issue might be or how to proceed with troubleshooting?

Oh, and just to further upset everything, this was all working fine and, as far as I'm aware, nothing has changed. It just stopped a little after midday, 9 days ago!

TIA

Paul

If I read correctly, everything used to work fine, now everything still works fine except from Windows 10.

From the timing, and from the fact that Win 10 had an update about 9 days ago, I’d guess it has something to do with the patch. I’d check the Windows firewall settings to see if the defaults haven’t been helpfully changed for you by Microsoft.

Thanks for the response!

Makes sense, unfortunately I already tried turning the firewall off in Windows.

I should also point out that the Windows boxes can "see" the host machine and the router fine (and vice-versa).

Does anyone know if Micro$oft's latest update did anything else with networking that may explain this?

My gut is telling me that the syslog message on the host machine may well have something to do with it, but I haven't been able to find out what may be causing it. I've checked for duplicate MACs and all the other things I have managed to find concern specific brands of routers or machines with both wLAN and LAN enabled....

Paul

I’ve heard Windows 10 does something different with ipv6. Maybe the other machines (ie iphone) looks for the ipv4 address first, which uniquely identifies your VM vs host, but for some reason, ipv6 is considering them the same machine. That might explain your syslog entries. There are so many different ways to setup bridged networking on a vm, that I don’t know where to start. I assume your VM is getting a unique MAC address. Are you using virtio? VDE? I’d probably setup networking for the VM from scratch again. You probably know all about it, but this https://www.linux-kvm.org/page/Networking might be helpful.

Start by checking which box has this MAC address. But who knows if this message is related to your problem.

I guess the Windows PC has connectivity to the VM host, but not the VM? Is that so for all VMs that you run there?

Check the ARP cache on the PC. If it doesn't contain the VM host (which I would expect), the PC must broadcast an ARP request that should be visible on the VM host. Then use tcpdump on the VM host to see if ARP requests arrive, and Wireshark on the PC to see if there is any reaction from anywhere.

Do the same tracing for other boxes and compare.