Routing Configuration Guide, Cisco Catalyst SD-WAN Releases 17.x

PDF

Route leaking

Updated: February 6, 2026

Overview

Explains the mechanisms used to share common services across multiple segments through bidirectional route replication. Understanding how OMP administrative distances and inter-service leaking influence path selection allows you to maintain granular control over your network segmentation and connectivity.

Route leaking is a mechanism that enables network segmentation using VPNs and allows sharing of common services that multiple VPNs need to access.

Route Leaking Between Global VRF and Service VPNs

Route leaking between the global or default VRF (transport VPN) and service VPNs allows you to share common services that multiple VPNs need to access. With this feature, routes are replicated through bidirectional route leaking between the global VRF (also known as transport VPN) and service VPNs. Route leaking between VRFs is done using Routing Information Base (RIB).

To leak routes to the routing neighbors, redistribute the leaked routes between the global VRF and service VPNs.

Note

In the context of Cisco Catalyst SD-WAN, the terms VRF and VPN are used interchangeably. Although Cisco IOS XE Catalyst SD-WAN devices use VRFs for segmentation and network isolation, the VPN feature template is used to configure them using Cisco SD-WAN Manager. When you use Cisco SD-WAN Manager to configure VPNs for Cisco IOS XE Catalyst SD-WAN devices, Cisco SD-WAN Manager automatically converts the VPN configuration to VRF configuration.

To leak routes to the routing neighbors, redistribute the leaked routes between the global VRF and service VPNs.

OMP administrative distance for leaked routes

You can configure the Cisco SD-WAN Overlay Management Protocol (OMP) administrative distance to a lower value that sets the OMP routes as the preferred and primary route over any leaked routes in a branch-to-branch routing scenario.

Ensure that you configure the OMP administrative distance on Cisco IOS XE Catalyst SD-WAN devices based on the following points:

  • If you configure the OMP administrative distance at both the global VRF and service VRF level, the VRF-level configuration overrides the global VRF-level configuration.

  • If you configure the service VRF with a lower administrative distance than the global VRF, then except the service VRF, all the remaining VRFs take the value of the administrative distance from the global VRF.

To configure the OMP administrative distance using Cisco SD-WAN Manager, see Configure Basic VPN Parameters and Configure OMP using SD-WAN Manager templates.

To configure the OMP administrative distance using the CLI, see the Configure OMP Administrative Distance section in Configure OMP Using the CLI.

Inter-Service VPN route leaking

The Inter-Service VPN Route Leaking feature provides the ability to leak selective routes between service VPNs back to the originating device on the same site.

To resolve routing-scalability challenges introduced when you use Cisco Catalyst SD-WAN Control Componentss, you can leak routes between the VPNs at the edge device.

To configure the inter-service VPN route leaking feature using Cisco SD-WAN Manager, see Configure Route Leaking Between Service VPNs.

To configure the inter-service VPNs route leaking feature using the CLI, see Configure Route Leaking Between Service VPNs Using the CLI.

Use VRRP tracker for leaked service VPNs

The Virtual Router Redundancy Protocol (VRRP) can track whether a leaked route is reachable. If tracked route is not reachable, VRRP changes the priority of the VRRP group. It can trigger a new primary router election. The VRRP tracker determines whether a route is reachable based on the existence of the route in the routing table of the routing instance that is included in the VRRP configuration.

To configure the VRRP tracker to track a leaked service VPN using Cisco SD-WAN Manager, see Configure VRRP for Cisco VPN Interface Ethernet template.

To configure the VRRP tracker to track any leaked service VPNs using the CLI, see Configure VRRP Tracker for Tracking Leaked Service VPNs Using the CLI.

Features of route leaking

Route leaking offers these features:

  • Routes between the global VRF and service VPNs can be leaked directly.

  • Multiple service VPNs can be leaked to the global VRF.

  • Multiple service VRFs leaking into the same service VRF is supported.

  • When routes are leaked or replicated between the global VRF and service VPNs, route properties such as metric, source VPN information, tags, administrative distance, and route origin are retained.

  • You can control leaked routes using route maps.

  • Route-maps can filter routes using match operations before leaking them.

  • The feature can be configured using both—Cisco SD-WAN Manager and CLI.


Use cases for route leaking

Route leaking provides solutions for various network scenarios, offering benefits such as service sharing, simplified migration, and enhanced network management.

Route leaking is applied in these scenarios:

  • Service Provider Central Services: This feature allows direct access to SP Central services under MPLS without duplicating them for each VPN. This approach makes accessing central services more efficient.

  • Migration: With route leaking, branches that have migrated to Cisco SD-WAN can directly access non-migrated branches bypassing the hub, thus providing improved application SLAs.

  • Centralized Network Management: You can manage the control plane and service-side equipment through the underlay.

  • Retailer Requirements for PCI compliance: Route leaking for service VPNs is used where the VPN traffic goes through a zone-based firewall on the same branch router while being PCI compliant.


How route preference is determined

When a route is replicated or leaked between the global VRF and service VPNs, a specific set of rules determines which route is preferred. These rules ensure consistent and predictable routing behavior within the network.

Route preference is determined by these rules:

  1. For a device that receives routes from two sources that both use the same source VRFs, and one of the routes is replicated, the non-replicated route is preferred.

  2. If the first rule does not apply, these rules determine route preference in this order:

    1. Prefer the route with smaller administrative distance.

    2. Prefer the route with smaller default administrative distance.

    3. Prefer a non-replicated route over a replicated route.

    4. Compare original VRF names. Prefer the route with the lexicographically smaller VRF name.

    5. Compare original subaddress families. Prefer unicast routing over multicast routing.

    6. Prefer the oldest route.