Skip to main content
  • Place orders quickly and easily
  • View orders and track your shipping status
  • Enjoy members-only rewards and discounts
  • Create and access a list of your products
  • Manage your Dell EMC sites, products, and product-level contacts using Company Administration.

Dell SmartFabric OS10 User Guide Release 10.5.3

PDF

EVPN local route advertisement

BGP EVPN running on each VTEP listens to local overlay information, encodes them as BGP EVPN routes, and injects them into BGP to be advertised to remote VTEPs. This section describes EVPN local route advertisement functionality when the VTEPs are set up as VLT pairs.

Overlay access MAC learn or age or flush or clear

For all MAC learn or age or flush or clear operations on local ports in the overlay, there is no change in the regular VLT functionality as defined for static Virtual-networks.

Meaning, when VLT node A learns a MAC address in the virtual-network from the data-plane on its access port, it synchronizes this MAC address with VLT to its VLT peer node B.

Similarly, in case of MAC age or clear or flush operations for MACs on VLT ports, the MAC is removed from both VLT peers only if it has aged on both nodes. Also, for MACs on orphan ports the MAC is immediately removed on both VLT peers.

The following topology diagram depicts the overlay access MAC learn or age or clear process:

overlay-access-mac-learn

Identical EVPN routes by both VLT peer nodes

Since each VLT node is a separate BGP router, both VLT nodes individually advertise EVPN route for a MAC irrespective of whether it is locally learnt or is synchronized by VLT from the peer. However, since the remote VTEPs see the VLT pair as a single logical VTEP, the remote VTEPs need to see only one copy of this route.

This behavior is achieved by ensuring that the EVPN routes advertised by both VLT peer nodes have the same NLRI and Next-hop. The Next-hop is always same, since it is the common NVE source IP address that is shared by both VLT peers.

The NLRI varies according to Route type.

Route Type 2 - MAC or IP

  • Advertised for each MAC learnt in each Virtual-network (EVI).
  • The NLRI consists of RD, MAC, IP, Ethernet tag ID, or MPLS Label1 (VNI). Ethernet tag ID is always 0 in this case (1-1 VNI to EVI).
  • Except RD, all the other parameters advertised by both VLT peers are the same.

Route Type 3 - Inclusive multicast route

  • Advertised for each Virtual-network (EVI) that has a VNI associated with it.
  • The NLRI consists of RD, Ethernet tag ID, or IP of VTEP. Ethernet tag ID is always 0 in this case (1-1 VNI to EVI).
  • Except RD, all the other parameters advertised by both VLT peers are the same.

RD is the Route distinguisher that is used to ensure that BGP maintains identical route information belonging to different nodes that are separate from each other.

For EVPN we use Type 1 RD, which comprises of IP address to identify the VTEP node. EVI to identify each Virtual-network or VNI.

In order to have identical RDs between VLT peers in all route types:

  • The common NVE source IP is used as the IP address in RD.
  • Auto-EVI mode -
    • The Virtual-network ID is used as the EVI.
    • The Virtual-network ID is explicitly configured by user on both VLT peers and is already mandated to be identical for a given VNI.
  • Manual-EVI mode -
    • If the RD is auto, then the EVI to VNI mapping that you configure needs to be identical on both VLT peers.
    • If the RD is manual, then the RD to VNI mapping that you configure needs to be identical on both VLT peers.

Rate this content

Accurate
Useful
Easy to understand
Was this article helpful?
0/3000 characters
  Please provide ratings (1-5 stars).
  Please provide ratings (1-5 stars).
  Please provide ratings (1-5 stars).
  Please select whether the article was helpful or not.
  Comments cannot contain these special characters: <>()\