Same hardware MAC showing up for 2 physical interfaces.

[McGee]

Hi All,

I'm a fairly new Proxmox user, coming over from ESXi in the last few weeks.
I'm enjoying the experience so far, but I've run into an issue that I'm a bit stumped on and can't seem to find a similar case on the Google's to try use for troubleshooting.

My installation is on a Minisforums MS-01 mini pc, with two SFPs and two copper ethernet connections.

I've found dmesg to be full of the duplicate packets received with a same mac;

[20816.681998] vmbr0: received packet on enp89s0 with own address as source address (addr:58:47:ca:76:76:fa, vlan:1)

As I look more closely at the NICs I notice what looks to be the two SFPs are sharing the same MAC, it is also replicated on the bond which I understand is normal.
It's also been replicated onto a virtual bridge, unsure if this is normal behavior or not.

I'm not sure if anyone has seen this before, or if Minisforums are faking the hardware addresses potentially, which would be a shame as I was under the understanding it was all intel based chipsets.
Please let me know what information I can supply to try and debug this.

it's very possible I've set something up wrong - but only one way to learn these things!

Cheers

root@proxmox-ms01:~# ifconfig
bond0: flags=5187<UP,BROADCAST,RUNNING,MASTER,MULTICAST> mtu 1500
ether 58:47:ca:76:76:f7 txqueuelen 1000 (Ethernet)
RX packets 1518381 bytes 325260918 (310.1 MiB)
RX errors 0 dropped 1 overruns 0 frame 0
TX packets 3518137 bytes 4940600673 (4.6 GiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

enp2s0f0np0: flags=6211<UP,BROADCAST,RUNNING,SLAVE,MULTICAST> mtu 1500
ether 58:47:ca:76:76:f7 txqueuelen 1000 (Ethernet)
RX packets 1059824 bytes 249938368 (238.3 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 3468122 bytes 4933581105 (4.5 GiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

enp2s0f1np1: flags=6211<UP,BROADCAST,RUNNING,SLAVE,MULTICAST> mtu 1500
ether 58:47:ca:76:76:f7 txqueuelen 1000 (Ethernet)
RX packets 458557 bytes 75322550 (71.8 MiB)
RX errors 0 dropped 1 overruns 0 frame 0
TX packets 50015 bytes 7019568 (6.6 MiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

vmbr2: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 172.16.60.252 netmask 255.255.255.0 broadcast 0.0.0.0
inet6 fe80::5a47:caff:fe76:76f7 prefixlen 64 scopeid 0x20<link>
ether 58:47:ca:76:76:f7 txqueuelen 1000 (Ethernet)
RX packets 920516 bytes 240015371 (228.8 MiB)
RX errors 0 dropped 67 overruns 0 frame 0
TX packets 24 bytes 3156 (3.0 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

root@proxmox-ms01:~# brctl show vmbr2
bridge name bridge id STP enabled interfaces
vmbr2 8000.5847ca7676f7 no bond0
tap201i0
veth100i0
veth101i0
veth102i0
veth103i0
veth104i0
veth105i0
veth106i0
veth107i0
veth109i0
veth115i0
veth118i0

 

[shanreich]

The bridge having the same MAC as the enslaved IF is to be expected, for the slaves of the bond as well (depending on the bond mode). It seems like you have another NIC (enp89s0) where packets originating from bond0 are received. This points more to an issue with your network setup/topology rather than an issue with Proxmox itself. How does the configuration for enp89s0 look like and how are the NICs hooked up?

Which bond mode are you using?

 

[McGee]

I suspect you are correct, I've had some unusual traffic flows occurring but haven't been able to track down the potential culprit.

Bond0 is using LACP 802.3ad.
Physically the bond0 is terminated into a unifi switch using an access vlan for the server network.
enp89s0 is also connected to the same switch on the same access vlan.

enp89s0 has the management interface and gateway attached to it that was used as the original connection when I first built the box.
I'm trying to move away from this and have bond0 as the primary interface, but have had issues with connectivity as soon as I drop the link connected to enp89s0, even from within the same subnet.

I figure I might have introduced some form of routing loop with this configuration accidentally.

Config for enp89s0 below;

iface enp89s0 inet manual
auto vmbr0


iface vmbr0 inet static
address 172.16.60.72/24
gateway 172.16.60.254
bridge-ports enp89s0
bridge-stp off
bridge-fd 0
bridge-vlan-aware yes
bridge-vids 2-4094




auto vlan55
iface vlan55 inet static
address 172.16.55.72/24
vlan-raw-device enp89s0

 

[McGee]

I managed to get this sorted, I was distracted by the double MAC which seems to be a null issue.
It turns out the Intel AMT interface was getting a DHCP lease different to the pve host which was causing the issue.
I've set it to be the same as the pve host and all is looking good.

Can now mark this as resolved!