Configuring VPLS Multihoming
VPLS multihoming allows you to connect a customer site to multiple PE routers to provide redundant connectivity while preventing the formation of Layer 2 loops in the service provider’s network. A VPLS site multihomed to two or more PE routers provides redundant connectivity in the event of a PE router-to-CE device link failure or the failure of a PE router. For more information about VPLS multihoming, see VPLS Multihoming Overview.
If you want to enable multihoming for a VPLS routing instance, you cannot also enable LDP signaling. You can only enable BGP signaling.
In the VPLS documentation, the word router in terms such as PE router is used to refer to any device that provides routing functions.
The following sections describe how to configure VPLS multihoming. Some information is also provided on single-homed site configuration versus multihomed site configuration.
VPLS Multihomed Site Configuration
The following describes the requirements for a VPLS multihomed site configuration:
Assign the same site ID on all PE routers connected to the same CE devices.
When two PE routers use the same site ID, VPLS assumes multihoming behavior by default. The site preference value is used to signal the primary and backup PE router. In such cases, when multihoming is explicitly configured using the
multi-homingstatement , it is only used for tracking the BGP peer status, such as to prevent isolation of the PE router from the core or split brain. There are scenarios, however that require preventing of BGP peer tracking, such as in a two-PE-router topology. In such cases, multihoming should not be explicitly configured as it can break node redundancy.When identical site IDs are used without configuring multihoming, a collision log message is generated at each signaling:
RPD_L2VPN_SITE_COLLISION: Same site ID 2 configured on remote PE (RD 8.8.8.1:1013:) and local PE in VPN 1013 (non-multihomed site 2). This is expected behavior.Assign unique route distinguishers for each multihomed PE router.
Reference all interfaces assigned to the multihomed VPLS site on each PE router. Only one of these interfaces is used to send and receive traffic for this site at a time.
Either designate a primary interface or allow the router to select the interface to be used as the primary interface.
If the router selects the interface, the interface used to connect the PE router to the site depends on the order in which interfaces are listed in the PE router’s configuration. The first operational interface in the set of configured interfaces is chosen to be the designated interface. If this interface fails, the next interface in the list is selected to send and receive traffic for the site.
The following configuration shows the statements you need to configure to enable VPLS multihoming:
[edit routing-instances routing-instance-name] instance-type vpls; interface interface-name; interface interface-name; protocols vpls { site site-name { active-interface { any; primary interface-name; } interface interface-name; interface interface-name; multi-homing; site-identifier number; } } route-distinguisher (as-number:id | ip-address:id);
If you add a direct connection between CE devices that are multihomed to the same VPLS site on different PE routers, the traffic can loop and a loss of connectivity might occur. We do not recommend this topology.
Most of these statements are explained in more detail in the rest of this chapter. The following sections explain how to configure the statements that are specific to VPLS multihoming:
Specifying an Interface as the Active Interface
You need to specify one of the interfaces for the multihomed site as the primary interface. If there are multiple interfaces, the remaining interfaces are activated only when the primary interface goes down. If no active interfaces are configured at the site level, all traffic for a VPLS site travels through a single, non-multihomed PE router.
You must configure one of the following options for the active-interface statement:
any—One configured interface is randomly designated as the active interface for the VPLS site.primary—Specify the name of the multihomed interface to be used as the primary interface by the VPLS site.
To specify a multihomed interface as the primary interface for
the VPLS site, include the active-interface statement:
active-interface { any; primary interface-name; }
You can include this statement at the following hierarchy levels:
[edit routing-instances routing-instance-name protocols vpls site site-name][edit logical-systems logical-system-name routing-instances routing-instance-name protocols vpls site site-name]
Configuring Multihoming on the PE Router
When a CE device is connected to the same VPLS site
on more than one PE router, including the multi-homing statement
on all associated PE routers results in tracking of BGP peers. If
no BGP peer is available, VPLS deactivates all active interfaces for
a site. When two PE routers use the same site ID, VPLS assumes multihoming
behavior by default. The site preference value is used to signal the
primary and backup PE router. In such cases, when multihoming is explicitly
configured using the multi-homing statement , it is only
used for tracking the BGP peer status, such as to prevent isolation
of the PE router from the core or split brain.
When identical site IDs are used without configuring multihoming,
a collision log message is generated at each signaling: RPD_L2VPN_SITE_COLLISION:
Same site ID 2 configured on remote PE (RD 8.8.8.1:1013:) and local
PE in VPN 1013 (non-multihomed site 2). This is expected behavior.
VPLS Single-Homed Site Configuration
All VPLS single-homed sites are connected to the same default VE device. All interfaces in a VPLS routing instance that are not configured as part of a multihomed site are assumed to be single-homed to the default VE device.