LACP with vlan for management nic

Harvester
1

so, i have 2x 10Gig Nics and i enabled lacp tagging multiple Vlans on the switch side. Now when i try to install harvester at configure network step. I pick both nics and fill in the vlan id with ip set to static. Like this attached screenshot:

Screenshot 2024-08-30 at 16.01.12
Screenshot 2024-08-30 at 16.01.121890×1050 123 KB

The installer fails the network config. If i remove the vlan id, the installer goes further but then i can’t ping the cluster management ip. So then Esc back to network config page. Add the vlan id, this time installer goes further and i can then ping the cluster ip and join the node to the cluster.

The problem is that once the node is rebooted. The node loses its network. Upon further checking i can see mgmt-br & mgmt-bo are up but mgmt-br.8 is down:

mgmt-br: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
mgmt-bo: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue master mgmt-br state UP group default qlen 1000

mgmt-br.8@mgmt-br: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state down group default qlen 1000

I can bring up mgmt-br.8 up with ifup and the node back up and all is good.

Looking further in the /oem/90_custom.yaml. I can see the mgmt-br.8 definition:

  • path: /etc/sysconfig/network/ifcfg-mgmt-br.8
    permissions: 384
    owner: 0
    group: 0
    content: |+
    STARTMODE=‘onboot’
    ETHERDEVICE=‘mgmt-br’
    VLAN_PROTOCOL=‘ieee802-1Q’
    VLAN_ID=8

             BOOTPROTO='static'

         IPADDR=10.7.96.92
         NETMASK=255.255.255.0


         DHCLIENT_SET_DEFAULT_ROUTE='yes'

So, can someone tell how can i make the mgmt-br.8 bridge comeup on each reboot rather than doing it manually each time or am i doing something wrong here?