Configuring Public Network IPv4 over SRv6 TE Policy

This section describes how to configure public network IPv4 over SRv6 TE Policy.

Usage Scenario

Public network IPv4 over SRv6 TE Policy uses SRv6 TE Policies on a public network to carry public network IPv4 services. The key implementation of public network IPv4 over SRv6 TE Policy includes establishing SRv6 TE Policies, advertising BGP routes, and forwarding data. As shown in Figure 1, PE1 and PE2 communicate through an IPv6 public network. An SRv6 TE Policy can be established on the IPv6 public network to carry public network IPv4 services.

Figure 1 Public network IPv4 over SRv6 TE Policy networking

Pre-configuration Tasks

Before configuring public network IPv4 over SRv6 TE Policy, complete the following tasks:

  • Configure a link layer protocol.

  • Configure network-layer addresses for interfaces to ensure that neighboring devices are reachable at the network layer.

Procedure

  1. Configure IPv6 IS-IS on the PEs and P. For configuration details, see Configuring Basic IPv6 IS-IS Functions.
  2. Establish an EBGP peer relationship between PE1 and DeviceA and another one between PE2 and DeviceB.
    1. Run system-view

      The system view is displayed.

    2. Run bgp as-number

      The BGP view is displayed.

    3. Run router-id ipv4-address

      A BGP router ID is configured.

    4. Run peer ipv4-address as-number as-number

      A peer address and the number of the AS where the peer resides are specified.

    5. Run commit

      The configuration is committed.

    6. Run quit

      Exit the BGP view.

  3. Establish an IBGP peer relationship between the PEs.
    1. Run bgp as-number

      The BGP view is displayed.

    2. Run peer ipv6-address as-number as-number

      The address of a peer and the number of the AS where the peer resides are specified.

    3. Run peer ipv6-address connect-interface interface-type interface-number

      The source interface and address used to establish a TCP connection are specified.

    4. Run ipv4-family unicast

      The BGP-IPv4 unicast address family view is displayed.

    5. Run peer ipv6-address enable

      The device is enabled to exchange routing information with the specified IPv6 peer.

    6. Run commit

      The configuration is committed.

    7. Run quit

      Exit the BGP-IPv4 unicast address family view.

    8. Run quit

      Exit the BGP view.

  4. Configure basic SRv6 functions.
    1. Run segment-routing ipv6

      SRv6 is enabled, and the SRv6 view is displayed.

    2. Run encapsulation source-address ipv6-address [ ip-ttl ttl-value ]

      A source address is specified for SRv6 encapsulation.

    3. Run locator locator-name [ ipv6-prefix ipv6-address prefix-length [ static static-length | args args-length ] * ]

      An SRv6 locator is configured.

    4. (Optional) Run opcode func-opcode end-dt4

      A static SID operation code (opcode) is configured.

      An End.DT4 SID can be either dynamically allocated through BGP or manually configured. If a dynamically allocated SID and a manually configured SID both exist, the latter takes effect. If you want to run the segment-routing ipv6 locator locator-name command to enable dynamic End.DT4 SID allocation through BGP, you can skip this step

    5. Run quit

      Exit the SRv6 locator view.

    6. Run quit

      Exit the SRv6 view.

  5. Enable IS-IS SRv6.
    1. Run isis [ process-id ]

      The IS-IS view is displayed.

    2. Run ipv6 enable topology ipv6

      The IPv6 capability is enabled for the IS-IS process in the IPv6 topology.

    3. Run segment-routing ipv6 locator locator-name [ auto-sid-disable ]

      IS-IS SRv6 is enabled.

    4. Run quit

      Exit the IS-IS view.

  6. Configure each PE to carry color and SID attributes in public network routes to be advertised.
    1. Run route-policy route-policy-name { permit | deny } node node

      A route-policy with a specified node is created, and the route-policy view is displayed.

    2. Run apply extcommunity color color

      A color extended community attribute is set.

    3. Run quit

      Exit the route-policy view.

    4. Run bgp as-number

      The BGP view is displayed.

    5. Run ipv4-family unicast

      The BGP-IPv4 unicast address family view is displayed.

    6. Run peer ipv6-address route-policy route-policy-name export

      The export route-policy is applied to routes to be advertised to the specified peer so that the color attribute can be carried in these routes.

    7. Run peer ipv6-address advertise-ext-community

      The device is enabled to advertise extended community attributes to the specified peer.

    8. Run peer ipv6-address prefix-sid

      The device is enabled to exchange prefix SID information with the specified peer.

    9. Run segment-routing ipv6 locator locator-name

      The device is enabled to carry SIDs in public network routes.

    10. (Optional) Run segment-routing ipv6 apply-sid per-nexthop pop-go

      The device is configured to allocate SIDs to IPv4 unicast routes received from public network IPv4 peers based on next hops.

    11. Run quit

      Exit the BGP-IPv4 unicast address family view.

    12. Run commit

      The configuration is committed.

  7. Configure an SRv6 TE Policy. For configuration details, see Configuring an SRv6 TE Policy (Manual Configuration) or Configuring an SRv6 TE Policy (Dynamic Delivery by a Controller).
  8. Configure public network routes to recurse to the SRv6 TE Policy.
    1. Run tunnel-policy policy-name

      A tunnel policy is created, and the tunnel policy view is displayed.

    2. Run tunnel select-seq ipv6 srv6-te-policy load-balance-number loadBalanceNumber

      The tunnel type to be selected and the maximum number of tunnels that can be used for load balancing are configured.

    3. Run quit

      Exit the tunnel policy view.

    4. Run tunnel-selector tunnel-selector-name { permit | deny } node node

      A tunnel selector is created, and the tunnel selector view is displayed.

    5. Run apply tunnel-policy tunnel-policy-name

      The specified tunnel policy is applied to routes.

    6. Run quit

      Exit the tunnel selector view.

    7. Run bgp as-number

      The BGP view is displayed.

    8. Run ipv4-family unicast

      The BGP-IPv4 unicast address family view is displayed.

    9. Run unicast-route recursive-lookup tunnel-v6 tunnel-selector tunnel-selector-name

      The function to recurse public network routes to IPv6 tunnels is enabled.

    10. Run segment-routing ipv6 traffic-engineer

      The device is enabled to recurse routes to the SRv6 TE Policy based on the color and SID values carried in the routes.

    11. Run commit

      The configuration is committed.

Verifying the Configuration

After configuring public network IPv4 over SRv6 TE Policy, verify the configuration.

Copyright © Huawei Technologies Co., Ltd.
Copyright © Huawei Technologies Co., Ltd.
< Previous topic Next topic >