Service Manual

Table Of Contents
When VLT peer-routing is enabled, BFD sessions are not established if BFD packets reach the wrong VLT node due to LAG
hashing in TOR. The wrong VLT node must have ARP or IPv6 neighbor discovery entries for VLT peer node's IPv4 or IPv6
address in the forwarding table, to forward the BFD packet to the VLT peer. Since ARP or IPv6 neighbor discovery entries are
bound to expire, it is recommended to add static ARP or IPv6 neighbor discovery entries for the VLT peer node's IPv4 or IPv6
address, to forward the BFD packets to the correct VLT node.
Sample BFD configuration in VLT domain
The following shows the sample configuration of BFD implementation in VLT environment:
ToR
1. Enable BFD globally.
TOR(conf)# bfd enable
2. Configure a VLT peer LAG.
TOR(conf)#interface tengigabitethernet 1/1/1
TOR(conf-if-te-1/1/1)#no ip address
TOR(conf-if-te-1/1/1)#port-channel-protocol lacp
TOR(conf-if-te-1/1/1)#port-channel 10 mode active
TOR(conf-if-te-1/1/1)#no shutdown
TOR(conf)#interface tengigabitethernet 1/1/2
TOR(conf-if-te-1/1/2)#no ip address
TOR(conf-if-te-1/1/2)#port-channel-protocol lacp
TOR(conf-if-te-1/1/2)#port-channel 10 mode active
TOR(conf-if-te-1/1/2)#no shutdown
3. Configure the port channel for the VLT interconnect on a ToR.
TOR(conf)# interface port-channel 10
TOR(conf-if-po-111)# no ip address
TOR(conf-if-po-111)# switchport
TOR(conf-if-po-111)# no shutdown
VLT Proxy Gateway
1009