Hi All,
This question, and a lot more detail, is also over on askubuntu.com. I know several users from here also frequent askubuntu.com, but for those that don't...
I prefer to run my VM's on my LAN, and therefore want to to use bridged mode. I have never been able to make this work properly on my 20.04 Ubuntu test server. It works fine on my systemd/networkd Debian server and used to work fine on pre-netplan (pre systemd/networkd) Ubuntu servers.
Packets that should be sent onwards to the vnet NICs are not. They don't even appear on tcpdump on the host bridge. The packets just seem to get dropped from the link layer. The behaviour seems much like non-promiscuous mode, but I have done specific tests ensuring everything is in promiscuous mode with the same results.
I never run containers (lxc, lxd, docker) and do not have the br_netfilter module loaded, nor any iptables rules at all, and I want it that way.
I have tried pretty much everything from every reference I have found.
This question, and a lot more detail, is also over on askubuntu.com. I know several users from here also frequent askubuntu.com, but for those that don't...
I prefer to run my VM's on my LAN, and therefore want to to use bridged mode. I have never been able to make this work properly on my 20.04 Ubuntu test server. It works fine on my systemd/networkd Debian server and used to work fine on pre-netplan (pre systemd/networkd) Ubuntu servers.
Packets that should be sent onwards to the vnet NICs are not. They don't even appear on tcpdump on the host bridge. The packets just seem to get dropped from the link layer. The behaviour seems much like non-promiscuous mode, but I have done specific tests ensuring everything is in promiscuous mode with the same results.
I never run containers (lxc, lxd, docker) and do not have the br_netfilter module loaded, nor any iptables rules at all, and I want it that way.
I have tried pretty much everything from every reference I have found.