Multiple networks on guest ct - Proxmox 3.4

[jernej]

Hello,
I have a proxmox host configured with two network cards. eth0 is the primary card with 10.0.6.0/24 network, bridged to vmbr0. There is another phisicaly separate network connected to second NIC with 10.0.2.0/24 subnet bridged to vmbr1.
I have a container with some monitoring software installed and need to ping IPs from both networks. A added two network devices eth0 (linked to vmbr0) and eth1 (linked to vmbr1) on the virtual container and also added ip address from both networks (10.0.6.209 and 10.0.2.209).

On the hardware node configuration seems ok and i can ping hosts from both subnets. However from the virtual machine I am unable to ping hosts from the 10.0.2.0/24 network. It seems that all packets go to the first network card.
I would appreciate any input!

config of proxmox host: http://pastebin.com/8Yy9WK5S
config of the virtual host: http://pastebin.com/QGA8L3dg

 

[Richard]

Two "venet0" NICs do not work as expected - use veth instead!

 

[Brian Chojnowski]

I have this same issue. VM os is Ubuntu Server 18.04. Like the original poster's network config, I have 2 NIC cards on the host, eth1 & eth2 attached to vmbr0 & vmbr1 respectively. Each vmbr is configured with an ip address for the connected subnet and vmbr0 has the default gateway assigned. Inside the VM, I get "destination host unreachable" trying to ping any device on the network connected to eth2/vmbr1.

 

[Brian Chojnowski]

let's say I have subnets 10.0.1.0 and 10.0.2.0. eno1 and eno2 are physical NICs on the proxmox host, and ens20 and ens21 are interfaces in the VM. my config then looks like:
subnet 10.0.1.x -> eno1 -> vmbr0 10.0.1.100/24 gw 10.0.1.1 -> ens20 -> 10.0.1.200/24
subnet 10.0.2.x -> eno2 -> vmbr1 10.0.2.100/24 -> ens21 -> 10.0.2.200/24
I can ping the proxmox web gui on 10.0.1.100 but not on 10.0.2.100 ?!?
From inside the VM, the route table looks good, just default route 10.0.1.1, and the two routes for each interface, but I can ping both 10.0.1.100 and 10.0.2.100. However, the vm cannot ping any other 10.0.2.x devices, nor can any of these other devices ping either the 10.0.2.100 web gui, nor the vm's 10.0.2.200 interface.
I have turned off the proxmon's firewall, thinking something there might be getting in the way, but no change.
Anyone have any idea's what I am doing wrong? From reading up on other people's configs associated with running a firewall in VM ( I do not have any firewall VMs, just a vanilla ubuntu server 18.04 install ) I believe this sort of configuration is expected to work.