ProxmoxVE: All IPs on a single interface

[Lime-Parallelogram]

Bug

If the virtual machine has multiple interfaces, all network configuration is applied to a single interface (the last one in the Proxmox list, presumably the highest PCI address?). This means that if you have 3 interfaces and set a static IP address for each, the last interface has 3 addresses assigned while the other two have no addresses.

Even if you only specify networking configuration for a single interface, it is applied to the last one no matter which interface it was supposed to relate to. I'm guessing after-burn is not matching up the MAC addresses when interface configuration is applied. See example below if that helps explain.

Operating System Version

Fedora CoreOS 42.20250929

Afterburn Version

Environment

What hardware/cloud provider/hypervisor is being used to run Afterburn?
Proxmox VE 8.4.5

Expected Behavior

Each virtual interface has a single static IP address assigned to it.

Actual Behavior

All virtual interfaces within the configured virtual machine are listed but the last one has all of the IP addresses while the other has no addresses.

Reproduction Steps

1. Create a new Fedora CoreOS Virtual Machine according to the documentation
2. Using the GUI attach a total of 3 different network adaptors
3. Use the CloudInit configuration page in the GUI to set static IP addresses for each of the interfaces (net0, net1 & net2)
4. Boot Fedora CoreOS for the first time, log in and run ip a

Other Information

Proxmox virtual machine config /etc/pve/qemu-server/30061.conf

boot: order=scsi0;ide2;net0
cicustom: vendor=coreos:snippets/nginx_proxy.ign
cores: 4
cpu: x86-64-v2-AES
ide0: local-zfs:vm-30061-cloudinit,media=cdrom
ide2: none,media=cdrom
ipconfig0: ip=192.168.0.5/24,gw=192.168.0.1
ipconfig1: ip=10.10.10.50/24,gw=10.10.10.1
ipconfig2: ip=10.10.20.50/24
memory: 4096
meta: creation-qemu=9.2.0,ctime=1761418103
name: yfossproxy
net0: virtio=BC:24:11:B6:83:88,bridge=vmbr0,firewall=1
net1: virtio=BC:24:11:F3:CB:69,bridge=pubnet,firewall=1
net2: virtio=BC:24:11:7E:7F:2E,bridge=hacknet,firewall=1
numa: 0
ostype: l26
scsi0: local-zfs:vm-30061-disk-0,size=10G
scsihw: virtio-scsi-single
serial0: socket
smbios1: uuid=f24962cf-f80d-42ac-9b8f-ee4ea46df519
sockets: 1
vga: serial0
vmgenid: a113423c-565f-4ef9-b1b8-fa04b5ed08d3

Configuration from cloudinit disk (i.e. /dev/sr0/network-config)

[proxymanager@yfossproxy tmp]$ cat mount/network-config 
version: 1
config:
    - type: physical
      name: eth0
      mac_address: 'bc:24:11:b6:83:88'
      subnets:
      - type: static
        address: '192.168.0.5'
        netmask: '255.255.255.0'
        gateway: '192.168.0.1'
    - type: physical
      name: eth1
      mac_address: 'bc:24:11:f3:cb:69'
      subnets:
      - type: static
        address: '10.10.10.50'
        netmask: '255.255.255.0'
        gateway: '10.10.10.1'
    - type: physical
      name: eth2
      mac_address: 'bc:24:11:7e:7f:2e'
      subnets:
      - type: static
        address: '10.10.20.50'
        netmask: '255.255.255.0'
    - type: nameserver
      address:
      - '194.168.4.100'
      - '194.168.8.100'

Output of ip a

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host noprefixroute 
       valid_lft forever preferred_lft forever
2: ens18: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether bc:24:11:b6:83:88 brd ff:ff:ff:ff:ff:ff
    altname enp0s18
    altname enxbc2411b68388
3: ens19: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether bc:24:11:f3:cb:69 brd ff:ff:ff:ff:ff:ff
    altname enp0s19
    altname enxbc2411f3cb69
4: ens20: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether bc:24:11:7e:7f:2e brd ff:ff:ff:ff:ff:ff
    altname enp0s20
    altname enxbc24117e7f2e
    inet 192.168.0.5/24 brd 192.168.0.255 scope global noprefixroute ens20
       valid_lft forever preferred_lft forever
    inet 10.10.10.50/24 brd 10.10.10.255 scope global noprefixroute ens20
       valid_lft forever preferred_lft forever
    inet 10.10.20.50/24 brd 10.10.20.255 scope global noprefixroute ens20
       valid_lft forever preferred_lft forever
    inet6 fe80::be24:11ff:fe7e:7f2e/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever

[vilhalmer]

I just ran into this as well. It doesn't help that proxmox generates eth* names with no way to override, but currently the proxmox metadata provider doesn't attempt to use the names or mac addresses at all in the generated kernel params:

afterburn/src/providers/proxmoxve/cloudconfig.rs

Line 187 in de23c02

fn rd_network_kargs(&self) -> Result<Option<String>> {

Rather than pass through proxmox's likely-incorrect interface names, I think it would be possible for afterburn to look up the correct interface names using the mac addresses and stick them in the ip= arguments? I am not super familiar with the state of the system when it is running to know if the interfaces have gotten their predictable names yet.