Description

On all Junos and Junos Evolved platforms when MOFRR is enabled. Multicast traffic blackholing may be happened when MOFRR primary link went down for any reason and hence service impact. Ideally, Traffic should not be impacted as it could use MOFRR backup link

Symptoms

in some corner cases after multiple link flapping for MOFRR Primary link , RPD and Kernel may ran out of sync regarding RPF-NHs.

on the below graph at normal state , ae40 is Primary link and ae42 is backup link

    +-------------+   
    |             |            
    |            ae40----------
    |     DTU     |P           
    |             |            
    +-----ae42----+            
            | B                
            |                
            |                 
            | 

after ae40 (MOFRR primary link) went down , RPD and kernel ran in out of sync as shown below             

user@router> show krt multicast-composite-next-hop frr-rpf    
   Multicast Composite Nexthop Id: 993, Address: 0x9b129e4   
     References: 204, Kernel References: 52                   
     Member NHH: 0 (0xd86211c) type 4
        Next hop: xx.xx.xx.xx via ae42.0 weight 0x1, selected    >>>RPD maps to ae42.0
        Session Id: 0x0
Kernel's view of Nexthops :

% nhinfo -di 993
<...>
rewrite header length: 0
associated random information length: 64
hardware token: 0000000000
==========================================================
RPF info type =   1
==========================================================
Element       =   1
ifl           =       ae40.0     (333)        >>>Kernel maps to ae40.0
outer label   =   0
inner label   =   0
weight        =   1
flags         =   0x0
session id    =   0
<...>
from PFE's view:

SMPC16(router vty)# show nhd mc sr

MoFRR_IIF(type=9) ID=0:
   idx   total       ifl                       weight      flags
  -----  -----  -----------------------------------------------------
    993      1       333                            1(B)   0x00008000     >>>>>>>>>>>>>>  ifl 333 , ae40.0  <<<<<which is down
because of this out of sync issue , PFE install route with NHID 993 pointing to ae40 (which is already down) and starts dropping traffic.

Solution

For fixing the out of sync issue , customer need to confirm that backup RE has not the same issue and then graceful master RE switchover should fix the problem.
For software fix, customer need to have the fix of PR1793196
as a workaround , customer need to disable MOFRR 

Modification History

2024-04-09 : Article Created