Symptom

Using IPv6 FRR causes IPv6 packets to be sent with ethertype 0x0800 (IPv4) vs 0x86DD (IPv6). This causes the next hop device to improperly handle the packet thus causing traffic loss.

Conditions

Have ipv6 unicast fast-reroute per-prefix configured under the IGP.

Workaround

Not using FRR for IPv6 will not cause traffic loss.

Further Problem Description