veth¶

A usual way to connect VRF together is to use a veth interface. The veth interfaces are virtual Ethernet devices that are always created in interconnected pairs. They can act as tunnels between network namespaces.

veth interfaces are similar to xvrf interfaces, with the following differences:

the MAC address can be configured on veth interfaces
veth interfaces are not flagged NOARP, meaning that ARP or NDP resolution is done when sending an IP packet through it
veth interfaces support IP configuration

See also

xvrf interfaces.

Here is an example of configuration where veth interfaces connect two VRF.

vsr running config# / vrf vr1
vsr running vrf vr1# interface veth veth-to-vr2
vsr running veth veth-to-vr2#! link-interface veth-to-vr1 link-vrf vr2
vsr running veth veth-to-vr2#! ipv4 address 10.1.1.1/24
vsr running veth veth-to-vr2#! / vrf vr2
vsr running vrf vr2#! interface veth veth-to-vr1
vsr running veth veth-to-vr1#! link-interface veth-to-vr2 link-vrf vr1
vsr running veth veth-to-vr1# ipv4 address 10.1.1.2/24
vsr running veth veth-to-vr1# commit

A YANG condition ensures that the binding of veth interfaces is consistent: the veth interfaces of a given pair must bind each other.

A route can then be added in vr2 to reach a network 10.100.0.0/16 through vr1:

vsr running config# vrf vr2
vsr running vrf vr2# routing static
vsr running static# ipv4-route 10.100.0.0/16 next-hop 10.1.1.1
vsr running static# commit

Let’s fetch the veth state inside vr1 after committing this configuration:

vsr running config# show state vrf vr1 interface veth
veth veth-to-vr2
    mtu 1500
    promiscuous false
    enabled true
    ipv4
        address 10.1.1.1/24
        ..
    ipv6
        address fe80::687e:84ff:fed1:cc6/64
        ..
    oper-status UP
    counters
        in-octets 738
        in-unicast-pkts 7
        in-discards 0
        in-errors 0
        out-octets 738
        out-unicast-pkts 7
        out-discards 0
        out-errors 0
        ..
    ethernet
        mac-address 6a:7e:84:d1:0c:c6
        ..
    link-interface veth-to-vr1
    link-vrf vr2
    ..

The same configuration can be made using this NETCONF XML configuration.

vsr running config# show config xml absolute
<config xmlns="urn:6wind:vrouter">
  <vrf>
    <name>vr1</name>
    <interface xmlns="urn:6wind:vrouter/interface">
      <veth xmlns="urn:6wind:vrouter/veth">
        <name>veth-to-vr2</name>
        <enabled>true</enabled>
        <ipv4>
          <enabled>true</enabled>
          <address>
            <ip>10.1.1.1/24</ip>
          </address>
        </ipv4>
        <link-interface>veth-to-vr1</link-interface>
        <link-vrf>vr2</link-vrf>
        (...)
      </veth>
    </interface>
  </vrf>
  <vrf>
    <name>vr2</name>
    <interface xmlns="urn:6wind:vrouter/interface">
      <veth xmlns="urn:6wind:vrouter/veth">
        <name>veth-to-vr1</name>
        <enabled>true</enabled>
        <ipv4>
          <enabled>true</enabled>
          <address>
            <ip>10.1.1.2/24</ip>
          </address>
        </ipv4>
        <link-interface>veth-to-vr2</link-interface>
        <link-vrf>vr1</link-vrf>
        (...)
      </veth>
    </interface>
  </vrf>
</config>