SDN and VLANs

spekta23

New Member
Apr 30, 2025
3
1
3
Hi everyone,

Having some trouble and I am not sure if its just not possible (as my networking is not the best)

Background
3 node cluster each node has a 1GbE interface and a 10GbE SFP+ interface. Cluster works just fine.

Each cluster looks like this (network) and migration, storage, management is okay

YAML:
auto lo
iface lo inet loopback

auto eno1
iface eno1 inet manual

auto enp1s0f0
iface enp1s0f0 inet manual
        pre-up ethtool -G enp1s0f1 tx 4096 rx 8192
#10G Network

iface enp1s0f1 inet manual

auto enp1s0f0.202
iface enp1s0f0.202 inet static
        address 192.168.202.2/29
#Migration

auto eno1.201
iface eno1.201 inet static
        address 192.168.201.2/29
#Cluster

auto enp1s0f0.110
iface enp1s0f0.110 inet static
        address 192.168.110.2/24
#Storage Network

auto vmbr0
iface vmbr0 inet manual
        bridge-ports eno1
        bridge-stp off
        bridge-fd 0
        bridge-vlan-aware yes
        bridge-vids 2-4094
#Services Network

auto vmbr0.203
iface vmbr0.203 inet static
        address 192.168.203.2/29
        gateway 192.168.203.1
#Management

auto vmbr1
iface vmbr1 inet manual
        bridge-ports enp1s0f0
        bridge-stp off
        bridge-fd 0
        bridge-vlan-aware yes
        bridge-vids 2-4094
#Storage

source /etc/network/interfaces.d/*

I have then defined 3 SDN networks (all are running okay or at least appear to be)

YAML:
root@pve01:/etc/pve/sdn# cat subnets.cfg
subnet: internal-192.168.220.0-24
        vnet internal
        gateway 192.168.220.1
        snat 1

subnet: k8s-192.168.150.0-24
        vnet k8s
        gateway 192.168.150.1
        snat 1

subnet: storage-192.168.110.0-24
        vnet storage
        gateway 192.168.110.1

root@pve01:/etc/pve/sdn# cat zones.cfg
vlan: internal
        bridge vmbr0
        ipam pve

vlan: k8s
        bridge vmbr0
        ipam pve

vlan: storage
        bridge vmbr1
        ipam pve

root@pve01:/etc/pve/sdn# cat vnets.cfg
vnet: internal
        zone internal
        alias internal-services
        tag 220

vnet: k8s
        zone k8s
        alias Kubernetes Services
        tag 150

vnet: storage
        zone storage
        alias Storage Network for Services
        tag 110

And finally on the host where eth0 is the k8s and enp6s19 is the storage (10GbE)

YAML:
network:
  version: 2
  ethernets:
    eth0:
      match:
        macaddress: "02:f4:25:15:27:0b"
      addresses:
      - "192.168.150.2/24"
      nameservers:
        addresses:
        - 192.168.150.1
        search:
        - xxx.net
      set-name: "eth0"
      routes:
      - to: "default"
        via: 192.168.150.1
      - to: 192.168.150.0/24
        via: 192.168.150.1
        table: 101
      routing-policy:
        - from: 192.168.150.0/24
          table: 101
    enp6s19:
      addresses:
      - "192.168.110.110/24"
      nameservers:
        addresses:
        - 192.168.110.1
        search:
        - xxx.net
      routes:
      - to: 192.168.110.0/24
        via: 192.168.110.1
        table: 102
      routing-policy:
      - from: 192.168.110.0/24
        table: 102

YAML:
root@rancher0101:/etc/netplan# ip route list
default via 192.168.150.1 dev eth0 proto static
192.168.110.0/24 dev enp6s19 proto kernel scope link src 192.168.110.110
192.168.150.0/24 dev eth0 proto kernel scope link src 192.168.150.2

I can route anywhere except VLAN 110 (Destination unreachable).

On the PVE host no issues on 110. So the question becomes can I actually have a VLAN on the base PVE 110 assigned and working and then re-use that same VLAN in SDN to pass into hosts to utilise the storage network (VLAN 110) to gain access to the 10GbE for NFS mounts etc?

Have I fundamentally misunderstood something, if so pleae let me know where have i gone wrong. I feel like I have tried everything.

I was also able to confirm on the Ubiquiti UDM side that it can see the device and any IP changes against the MAC address so can only assume I have made an error in config.
 
Thank you spirit!

It seems obvious when you think about it.

So I have gone with this configuration and then just VLAN tag to 110 and static IP in the guest VM. Seems to work just fine.

YAML:
auto lo
iface lo inet loopback

auto eno1
iface eno1 inet manual

auto enp1s0f0
iface enp1s0f0 inet manual
        pre-up ethtool -G enp1s0f1 tx 4096 rx 8192
#10G Network

iface enp1s0f1 inet manual

auto enp1s0f0.202
iface enp1s0f0.202 inet static
        address 192.168.202.2/29
#Migration

auto eno1.201
iface eno1.201 inet static
        address 192.168.201.2/29
#Cluster

auto vmbr0
iface vmbr0 inet manual
        bridge-ports eno1
        bridge-stp off
        bridge-fd 0
        bridge-vlan-aware yes
        bridge-vids 2-4094
#Services Network

auto vmbr0.203
iface vmbr0.203 inet static
        address 192.168.203.2/29
        gateway 192.168.203.1
#Management

auto vmbr1
iface vmbr1 inet manual
        bridge-ports enp1s0f0
        bridge-stp off
        bridge-fd 0
        bridge-vlan-aware yes
        bridge-vids 2-4094
#10G Guest

auto vmbr1.110
iface vmbr1.110 inet static
        address 192.168.110.2/24
#Storage Network
 
  • Like
Reactions: spirit