Building BGP EVPN VXLAN datacenter fabric – Part 5 (NVE peering)

Quick Intro

The next building block of our DC network will be the NVE peering setup, which will use a prepared BGP network.
The NVE interface is the logical representation of a VTEP on a switch. It is responsible for handling the VXLAN encapsulation and decapsulation of Ethernet frames.
NVEs (or VTEPs) peer with each other using BGP EVPN to exchange MAC-to-IP mappings and other relevant VXLAN forwarding information.
So, following the information provided, I’ll configure NVE interfaces on each leaf switch.

Leaf switch configuration

feature nv overlay
!
interface nve1
  no shutdown
  host-reachability protocol bgp
  advertise virtual-rmac
  source-interface loopback0

Verification

Every leaf switch should have the following state of the NVE interface after the applied configuration, of course with the difference related to IP and mac-addresses

dc01-r01-leaf01# show nve interface nve 1 detail 
Interface: nve1, State: Up, encapsulation: VXLAN
 VPC Capability: VPC-VIP-Only [notified]
 Local Router MAC: 5003.0000.1b08
 Host Learning Mode: Control-Plane
 Source-Interface: loopback0 (primary: 10.255.255.3, secondary: 10.255.255.101)
 Source Interface State: Up
 Virtual RMAC Advertisement: Yes
 NVE Flags: 
 Interface Handle: 0x49000001
 Source Interface hold-down-time: 180
 Source Interface hold-up-time: 30
 Remaining hold-down time: 0 seconds
 Virtual Router MAC: 0200.0aff.ff65
 Interface state: nve-intf-add-complete

At this moment we’ll not be able to check any other outputs since I didn’t deploy any L2 or L3 services, so NVE peering is not up.

NVE Peering is up only when there is a learned EVPN route from the Peer and Peer IP is in RIB

Leave a comment