Rate and give feedback:  Feedback Received. Thank You!

Rate and give feedback: 

X

This document helped resolve my issue

Yes No

Additional Comments

800 characters remaining

May we contact you if necessary?

Name:  

E-mail: 

Enter a valid Email ID

Need product assistance? Contact Juniper Support

Submit

This site is protected by hCaptcha and its Privacy Policy and Terms of Service apply.

English  

English

Enabling BGP Path Selection for Layer 2 VPNs and VPLS

Layer 2 VPNs and VPLS share the same path selection process for determining the optimal path to reach all of the destinations shared within a single routing instance. For Layer 2 VPN and VPLS topologies that do not include multihomed PE routers, the path selection process is straightforward since there is just a single path from each PE router to each CE device. However, if multihoming is configured for one or more of the CE devices, the path selection process becomes more complex, since there can be two or more valid paths to reach each multihomed CE device.

Note: 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 Layer 2 VPN and VPLS path selection process uses the following path selection algorithms:

• On the Provider routers within the service providers network, the standard BGP path selection algorithm is used. Using the standard BGP path selection for Layer 2 VPN and VPLS routes allows service providers to leverage their existing Layer 3 VPN network infrastructure to also support Layer 2 VPNs and VPLS. The BGP path selection algorithm also helps to ensure that the service provider’s network behaves predictably with regard to Layer 2 VPN and VPLS path selection. This is particularly important in networks employing route reflectors and multihoming.

  When a Provider router receives multiple paths for the same destination prefix (for example, a multihomed CE device), one path is selected based on the BGP path selection algorithm and placed in the bgp.l2vpn.0 routing table and the appropriate instance.l2vpn.0 routing table. However, all of the available paths (including the backup paths for multihomed CE devices) are advertised to the intermediate BGP routers and the PE routers in the Layer 2 VPN or VPLS routing instances.

  For more information about the BGP path selection process, see Understanding BGP Path Selection.

• Once a PE router receives all of the available paths to each CE device, it runs the designated forwarder path selection algorithm to select the preferred path to reach each CE device, independently of the results of the earlier BGP path selection algorithm run on the Provider router. The VPLS designated forwarder algorithm uses the D-bit, preference, and PE router identifier to determine which of the valid paths to each CE device to use. The PE router might select a path to reach a CE device which is different from the path selected by the BGP-based Provider routers. In this scenario, the following is the expected behavior for traffic sent to the multihomed CE device:

  1. If the path selected by the remote PE router is available, traffic will traverse the network to the multihomed CE device using the remote PE router’s preferred path (again, ignoring the path selected by the BGP-based Provider routers).
  2. If the path selected by the remote PE router fails, the Provider routers switch the traffic destined for the multihomed CE device to the alternate path as soon as failure is detected. They then notify the remote PE router of the path failure. The remote PE router updates its routing table accordingly.
  For more information about the VPLS designated forwarder path selection algorithm, see VPLS Path Selection Process for PE Routers. This algorithm is also described in the Internet draft draft-ietf-l2vpn-vpls-multihoming-03.txt, BGP based Multi-homing in Virtual Private LAN Service.