Customer upgraded sbfd responder device. Replaced RE with new version and loaded saved configuration.
After that, customer found all sbfd sesstion to this responder are down.
user@mx> show spring-traffic-engineering lspTo State LSPname10.0.0.2 Down SPRING-TE-LSP
user@mx> show spring-traffic-engineering lsp
To State LSPname
10.0.0.2 Down SPRING-TE-LSP
From fpc shell commmand, there is no sbfd session.
SMPC0(mx vty)# show ppm adjacencies protocol bfd<<<< no sbfd session>>>>
SMPC0(mx vty)# show ppm adjacencies protocol bfd
<<<< no sbfd session>>>>
This issue can be reproduced follow below steps:
1.Delete all configuration, make sure there is "no lo0 configuration".
2.Configure lo0.100 and local-discriminator and then commit.
user@mx# set interfaces lo0 unit 100 family inet address 10.0.0.1/32 primaryuser@mx# set protocols bfd sbfd local-discriminator 20000001
user@mx# set interfaces lo0 unit 100 family inet address 10.0.0.1/32 primary
user@mx# set protocols bfd sbfd local-discriminator 20000001
Not having lo0.0 configuration for sbfd is not acceptable as sbfd responder run in distributed/inline mode, for bfd distribution configuring lo0.0 is must.
Refer below reference:
https://www.juniper.net/documentation/us/en/software/junos/high-availability/topics/topic-map/bfd.html#:~:text=CPU%20is%20congested.-,Distributed%20Configuration%20and%20Support,0%20and%20the%20appropriate%20family.&text=This%20is%20true%20for%20the,%2C%20and%20VPLS)%20(MPLS)
“For distributed BFD to work, you need to configure the lo0 interface with unit 0 and the appropriate family.”
Restart ppmd can recover this error status.
> restart ppmd immediately
Adding lo0.0 configuration can avoid this issue.