So, I have noticed some interesting behavior with proxmox's networking. Consider the below hosts network configuration:
So, as you can see, I have several linux virtual bridges that are tied to several unique, physical nics on the host. Now, the network associated with the bridge that is configured with the gateway (will call vmbr0 going forward) works as I would expect: the physical machines attached to that same network can see the virtual machines whose nics are linked to vmbr0 and vice versa. However, this is not the case if we look at vmbr1, As you can see, vmbr1 does not have its default gateway property configured since only one default gateway can be configured per host, and the default gateway property has already been set for vmbr0. However, for some reason, vms whose nics are configured to use the vmbr1 linux bridge are unable to see physical devices that are connected to the same network that vmbr1 bridges. Take a look at the below image:
Ok, so what I did was I ran a network scan on both the physical device that is connected to the same network that vmbr1 bridges and and on a vm whose nic is linked to vmbr1. I also ran an ipconfig on each device so that we can see each device's ip address. We know that these devices are connected to the same network as the mac address for their default gateway is identical. However, if we look at the results of the network scan, we can see that the network scan run on the vm is returning all of the vms that are attached to vmbr1, but notice how the it just so happens to skip right over the ip address associated with the phyisical machine? Even though it is online. And the physical machine's network scan couldn't see any of the vms. Why is this happening?
![2026-05-18 02_35_41-nirsoft (Ctrl+F12 - properties) [Control].png](https://forum.proxmox.com/data/attachments/97/97458-1ad0233505add71f7806813219d6eb68.jpg?hash=LWOjV2_RqJ "2026-05-18 02_35_41-nirsoft (Ctrl+F12 - properties) [Control].png")
So, as you can see, I have several linux virtual bridges that are tied to several unique, physical nics on the host. Now, the network associated with the bridge that is configured with the gateway (will call vmbr0 going forward) works as I would expect: the physical machines attached to that same network can see the virtual machines whose nics are linked to vmbr0 and vice versa. However, this is not the case if we look at vmbr1, As you can see, vmbr1 does not have its default gateway property configured since only one default gateway can be configured per host, and the default gateway property has already been set for vmbr0. However, for some reason, vms whose nics are configured to use the vmbr1 linux bridge are unable to see physical devices that are connected to the same network that vmbr1 bridges. Take a look at the below image:
Ok, so what I did was I ran a network scan on both the physical device that is connected to the same network that vmbr1 bridges and and on a vm whose nic is linked to vmbr1. I also ran an ipconfig on each device so that we can see each device's ip address. We know that these devices are connected to the same network as the mac address for their default gateway is identical. However, if we look at the results of the network scan, we can see that the network scan run on the vm is returning all of the vms that are attached to vmbr1, but notice how the it just so happens to skip right over the ip address associated with the phyisical machine? Even though it is online. And the physical machine's network scan couldn't see any of the vms. Why is this happening?
Last edited: