unconfigured interfaces part of a 802.3ad bond getting DHCP IP & being set as default route

[Richard Powell]

Hi,

I have the following configuration:

network:
  version: 2
  renderer: networkd

  ethernets:
    enp1s0:
      optional: true
    enp2s0:
      optional: true
    enp3s0:
      optional: true
    enp4s0:
      optional: true

  bridges:
    br0:
      interfaces: [enp2s0, enp4s0]
      addresses: [192.168.1.1/24]

  bonds:
    bond0:
      interfaces: [enp1s0, enp3s0]
      dhcp4: yes
      dhcp6: yes
      parameters:
        mode: 802.3ad
        mii-monitor-interval: 100ms
        lacp-rate: fast

The issue I'm having is that after doing a "netplan apply" or rebooting my machine the enp1s0 and enp2s0 interfaces that are part of bond0 are getting assigned IPs:

richard@server:/etc/netplan$ ifconfig
bond0: flags=5187<UP,BROADCAST,RUNNING,MASTER,MULTICAST> mtu 1500
        inet 67.170.73.230 netmask 255.255.252.0 broadcast 255.255.255.255
        inet6 fe80::f0e6:9eff:fe8e:f2c4 prefixlen 64 scopeid 0x20<link>
        inet6 2001:558:600a:31:6403:c1b5:aae4:9aba prefixlen 128 scopeid 0x0<global>
        ether f2:e6:9e:8e:f2:c4 txqueuelen 1000 (Ethernet)
        RX packets 439850 bytes 152153061 (152.1 MB)
        RX errors 0 dropped 0 overruns 0 frame 0
        TX packets 623764 bytes 705580476 (705.5 MB)
        TX errors 0 dropped 112 overruns 0 carrier 0 collisions 0

br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
        inet 192.168.1.1 netmask 255.255.255.0 broadcast 192.168.1.255
        inet6 fe80::280d:a8ff:fe45:7c23 prefixlen 64 scopeid 0x20<link>
        ether 2a:0d:a8:45:7c:23 txqueuelen 1000 (Ethernet)
        RX packets 574458 bytes 704661206 (704.6 MB)
        RX errors 0 dropped 1246 overruns 0 frame 0
        TX packets 416177 bytes 142338426 (142.3 MB)
        TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

enp1s0: flags=6211<UP,BROADCAST,RUNNING,SLAVE,MULTICAST> mtu 1500
        inet 67.170.73.230 netmask 255.255.252.0 broadcast 255.255.255.255
        ether f2:e6:9e:8e:f2:c4 txqueuelen 1000 (Ethernet)
        RX packets 265168 bytes 99714321 (99.7 MB)
        RX errors 0 dropped 0 overruns 0 frame 0
        TX packets 2251 bytes 138970 (138.9 KB)
        TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
        device memory 0xf7b00000-f7b1ffff

enp2s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
        ether 00:0e:c4:d4:45:12 txqueuelen 1000 (Ethernet)
        RX packets 597110 bytes 669525498 (669.5 MB)
        RX errors 0 dropped 0 overruns 0 frame 0
        TX packets 357068 bytes 66555302 (66.5 MB)
        TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
        device memory 0xf7a00000-f7a1ffff

enp3s0: flags=6211<UP,BROADCAST,RUNNING,SLAVE,MULTICAST> mtu 1500
        inet 67.170.73.230 netmask 255.255.252.0 broadcast 255.255.255.255
        ether f2:e6:9e:8e:f2:c4 txqueuelen 1000 (Ethernet)
        RX packets 174682 bytes 52438740 (52.4 MB)
        RX errors 0 dropped 0 overruns 0 frame 0
        TX packets 621513 bytes 705441506 (705.4 MB)
        TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
        device memory 0xf7900000-f791ffff

enp4s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
        ether 00:0e:c4:d4:45:14 txqueuelen 1000 (Ethernet)
        RX packets 107349 bytes 52876225 (52.8 MB)
        RX errors 0 dropped 0 overruns 0 frame 0
        TX packets 127306 bytes 83621171 (83.6 MB)
        TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
        device memory 0xf7800000-f781ffff

Even more problematic is the fact that those two interfaces are also being set as the default route ahead of bond0:

Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
0.0.0.0 67.170.72.1 0.0.0.0 UG 202 0 0 enp1s0
0.0.0.0 67.170.72.1 0.0.0.0 UG 204 0 0 enp3s0
0.0.0.0 67.170.72.1 0.0.0.0 UG 207 0 0 bond0
67.170.72.0 0.0.0.0 255.255.252.0 U 202 0 0 enp1s0
67.170.72.0 0.0.0.0 255.255.252.0 U 204 0 0 enp3s0
67.170.72.0 0.0.0.0 255.255.252.0 U 207 0 0 bond0
169.254.0.0 0.0.0.0 255.255.0.0 U 208 0 0 docker0
169.254.0.0 0.0.0.0 255.255.0.0 U 210 0 0 vethd01d043
169.254.0.0 0.0.0.0 255.255.0.0 U 212 0 0 vetha67b8e5
169.254.0.0 0.0.0.0 255.255.0.0 U 214 0 0 veth7dd8d67
169.254.0.0 0.0.0.0 255.255.0.0 U 216 0 0 vethfecc79a
172.17.0.0 0.0.0.0 255.255.0.0 U 0 0 0 docker0
192.168.1.0 0.0.0.0 255.255.255.0 U 0 0 0 br0

This really doesn't make sense to me since those should be "unconfigured" interfaces. I really want those interfaces to not show in my default routing table at all. Thanks for the help.

[Ryan Harper (raharper)]

I believe if you put:

dhcp4: no
dhcp6: no

under those optional interfaces, that should prevent netplan for asking to dhcp on those interfaces.

The docs indicate that dhcp4/dhcp6 are off by default, so there may still be a bug here
in that the composed devices (bridge/bonds) dhcp values are getting set on members.

[Launchpad Janitor (janitor)]

[Expired for netplan because there has been no activity for 60 days.]

[Launchpad Janitor (janitor)]

[Expired for netplan.io (Ubuntu) because there has been no activity for 60 days.]