Issues with MP-BGP and IPv6 Link-local peers using extended next-hop

Funny enough, that AS4242420207 is my buddy and I. We experience the same issue at times too and we run VyOS 1.5, it’s not every peer it’s only some.

We haven’t been able to figure out a cause either. We have noted rebooting fixes it, if we tcpdump on the wireguard interface it fixes it, etc.

Your configuration looks very similar to ours for you

 wireguard wg4242422227 {
     address 172.20.19.83/32
     address fe80::207/64
     description "DN42: AS4242422227 PUDUNET-SYD"
     ip {
         adjust-mss 1380
     }
     ipv6 {
         address {
             no-default-link-local
         }
         adjust-mss 1360
     }
     peer p2p {
         address x.x.x.x
         allowed-ips 0.0.0.0/0
         allowed-ips ::/0
         port 20207
         public-key xxx
     }
     port 52227
     private-key xxx
     vrf DN42
 }

 neighbor fe80::9898 {
     description PUDUNET-SYD
     interface {
         source-interface wg4242422227
     }
     peer-group DN42-PEER-MP-EXT-NH
     remote-as 4242422227
     strict-capability-match
     update-source fe80::207
 }
 peer-group DN42-PEER-MP-EXT-NH {
     address-family {
         ipv4-unicast {
             maximum-prefix 1000
             route-map {
                 export DN42-PEER-OUTBOUND
                 import DN42-PEER-INBOUND
             }
             soft-reconfiguration {
                 inbound
             }
         }
         ipv6-unicast {
             maximum-prefix 1000
             route-map {
                 export DN42-PEER-OUTBOUND
                 import DN42-PEER-INBOUND
             }
             soft-reconfiguration {
                 inbound
             }
         }
     }
     capability {
         extended-nexthop
     }
     disable-connected-check
 }

Not sure if it’s really VyOS related or FRRouting related either honestly.