Description of problem:

When configure MTU = 9000 on virtual controllers, all interfaces get the highest MTU from the underlying device configured, which differs from what is explained in [1]

[1] https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux_OpenStack_Platform/7/html/Director_Installation_and_Usage/sect-Configuring_Jumbo_Frames.html

How reproducible:
always

Steps to Reproduce:

we want InternalApiNetworkVlanID to stay with MTU 1500 :
        os_net_config:
          network_config:
            -
              type: ovs_bridge
              name: {get_input: bridge_name}
              mtu: 9000
              use_dhcp: false
              dns_servers: {get_param: DnsServers}
              addresses:
                -
                  ip_netmask:
                    list_join:
                      - '/'
                      - - {get_param: ControlPlaneIp}
                        - {get_param: ControlPlaneSubnetCidr}
              routes:
                -
                  ip_netmask: 169.254.169.254/32
                  next_hop: {get_param: EC2MetadataIp}
              members:
                -
                  type: interface
                  name: nic1
                  mtu: 9000
                  # force the MAC address of the bridge to this interface
                  primary: true
                -
                  type: vlan
                  vlan_id: {get_param: InternalApiNetworkVlanID}
                  addresses:
                  -
                    ip_netmask: {get_param: InternalApiIpSubnet}
                -
                  type: vlan
                  vlan_id: {get_param: StorageNetworkVlanID}
                  mtu: 9000
                  addresses:
                  -
                    ip_netmask: {get_param: StorageIpSubnet}
                -
                  type: vlan
                  vlan_id: {get_param: StorageMgmtNetworkVlanID}
                  mtu: 9000
                  addresses:
                  -
                    ip_netmask: {get_param: StorageMgmtIpSubnet}
                -
                  type: vlan
                  vlan_id: {get_param: TenantNetworkVlanID}
                  mtu: 9000
                  addresses:
                  -
                    ip_netmask: {get_param: TenantIpSubnet}

Also especially set mtu: 1500 for the InternalApiNetworkVlanID (VLAN 201) do not help as from : /usr/lib/python2.7/site-packages/os_net_config
    130         if base_opt.mtu != 1500:
    131             data += "MTU=%i\n" % base_opt.mtu

With this it is not possible to configure MTU 1500 for such a configuration.

On the controller we see the following in  os-apply-config :
Nov 18 09:32:38 localhost os-collect-config: ++ os-apply-config --key os_net_config --type raw --key-default ''
Nov 18 09:32:39 localhost os-collect-config: + NET_CONFIG='{"network_config": [{"dns_servers": ["8.8.8.8", "8.8.4.4"], "addresses": [{"ip_netmask": "192.0.2.12/24"}], "mtu": 9000, "members": [{"type": "interface", "name": "nic1", "primary": true, "mtu": 9000}, {"type": "vlan", "add
resses": [{"ip_netmask": "172.16.0.13/24"}], "vlan_id": 201, "mtu": 1500}, {"type": "vlan", "addresses": [{"ip_netmask": "172.18.0.11/24"}], "vlan_id": 202, "mtu": 9000}, {"type": "vlan", "addresses": [{"ip_netmask": "172.19.0.12/24"}], "vlan_id": 203, "mtu": 9000}, {"type": "vlan"
, "addresses": [{"ip_netmask": "172.17.0.11/24"}], "vlan_id": 204, "mtu": 9000}], "routes": [{"ip_netmask": "169.254.169.254/32", "next_hop": "192.0.2.1"}], "use_dhcp": false, "type": "ovs_bridge", "name": "br-ex"}, {"routes": [{"ip_netmask": "0.0.0.0/0", "next_hop": "192.168.122.1
"}], "type": "interface", "name": "nic2", "addresses": [{"ip_netmask": "192.168.122.51/24"}]}, {"use_dhcp": false, "defroute": false, "name": "nic3", "type": "interface"}]}'
Nov 18 09:32:39 localhost os-collect-config: + '[' -n '{"network_config": [{"dns_servers": ["8.8.8.8", "8.8.4.4"], "addresses": [{"ip_netmask": "192.0.2.12/24"}], "mtu": 9000, "members": [{"type": "interface", "name": "nic1", "primary": true, "mtu": 9000}, {"type": "vlan", "address
es": [{"ip_netmask": "172.16.0.13/24"}], "vlan_id": 201, "mtu": 1500}, {"type": "vlan", "addresses": [{"ip_netmask": "172.18.0.11/24"}], "vlan_id": 202, "mtu": 9000}, {"type": "vlan", "addresses": [{"ip_netmask": "172.19.0.12/24"}], "vlan_id": 203, "mtu": 9000}, {"type": "vlan", "a
ddresses": [{"ip_netmask": "172.17.0.11/24"}], "vlan_id": 204, "mtu": 9000}], "routes": [{"ip_netmask": "169.254.169.254/32", "next_hop": "192.0.2.1"}], "use_dhcp": false, "type": "ovs_bridge", "name": "br-ex"}, {"routes": [{"ip_netmask": "0.0.0.0/0", "next_hop": "192.168.122.1"}],
 "type": "interface", "name": "nic2", "addresses": [{"ip_netmask": "192.168.122.51/24"}]}, {"use_dhcp": false, "defroute": false, "name": "nic3", "type": "interface"}]}' ']'

Results in :

# cat /etc/sysconfig/network-scripts/ifcfg-vlan201 
# This file is autogenerated by os-net-config
DEVICE=vlan201
ONBOOT=yes
HOTPLUG=no
NM_CONTROLLED=no
DEVICETYPE=ovs
TYPE=OVSIntPort
OVS_BRIDGE=br-ex
OVS_OPTIONS="tag=201"
BOOTPROTO=static
IPADDR=172.16.0.13
NETMASK=255.255.255.0


Version-Release number of selected component (if applicable):
os-net-config-0.1.4-4.el7ost.noarch

Actual results:
all interfaces get MTU 9000 set

Expected results:
it is possible to configure MTU 1500 for VLAN interfaces even if the underlying interface has MTU 9000 configured