When RTG is enabled and any FPC member goes for a reboot, without any associated RTG member on the FPC which is going down, the switch starts to indefinitely incorrectly forward traffic in the backup member, which causes cascading issues in the network.

The problem can be seen after rebooting one member of the VC that has no link associated with RTG.

Topology:

DUT is a 4 members VC (0-3).

DUT FPC0---xe-0/0/48:0----xe-2/0/9 Peer-A
DUT FPC1---xe-1/0/48:0----xe-2/0/9 Peer-B (backup)

Before rebooting:

root@DUT# run show redundant-trunk-group      
Jul 18 16:20:49
Group      Interface   State       Time of last flap                      Flap 
name                                                                      count

rtg0       xe-0/0/48:0.0 Up/Pri/Act 2024-07-18 13:33:30 PDT (02:47:19 ago)     1
           xe-1/0/48:0.0 Up        2024-07-18 15:57:35 PDT (00:23:14 ago)     3

No ingress traffic in Peer-B:

root@Peer-B# run show spanning-tree statistics interface xe-2/0/9    
Jul 18 16:19:42

Interface     BPDUs       BPDUs        Next BPDU       TCs        Proposal    Agreement
              Sent        Received     Transmission    Tx/Rx      Tx/Rx       Tx/Rx    
xe-2/0/9             9           0             0           0/0         0/0         0/0  

root@Peer-B# run show spanning-tree interface xe-2/0/9 
Jul 18 16:20:07

Spanning tree interface parameters for VLAN 3

Interface                  Port ID    Designated         Designated         Port    State  Role
                                       port ID           bridge ID          Cost
xe-2/0/9                   128:535      128:535  16387.f01c2d2bedd0         2000    FWD    DESG

After reboot of fpc3 in the DUT VC:

No flaps in the RTG links:

root@DUT# run show redundant-trunk-group 
Jul 18 16:25:45
Group      Interface   State       Time of last flap                      Flap 
name                                                                      count

rtg0       xe-0/0/48:0.0 Up/Pri/Act 2024-07-18 13:33:30 PDT (02:52:15 ago)     1
           xe-1/0/48:0.0 Up        2024-07-18 15:57:35 PDT (00:28:10 ago)     3

We start receiving BPDUs in Peer-B.

root@Peer-B# run show spanning-tree interface xe-2/0/9    
Jul 18 16:22:12

Spanning tree interface parameters for VLAN 3

Interface                  Port ID    Designated         Designated         Port    State  Role
                                       port ID           bridge ID          Cost
xe-2/0/9                   128:535      128:538   8195.4c96149a6fd0         2000    BLK    ALT (Root-Prev)


root@Peer-B# run show spanning-tree statistics interface xe-2/0/9 
Jul 18 16:22:31

STP interface statistics for VLAN 3


Interface     BPDUs       BPDUs        Next BPDU       TCs        Proposal    Agreement
              Sent        Received     Transmission    Tx/Rx      Tx/Rx       Tx/Rx    
xe-2/0/9            85          15             0           0/0         0/0         0/0  

To recover from this state, you can deactivate and reactivate the RTG links.

set interface xe-0/0/48:0 disable

The permanent fix will provided via:

https://prsearch.juniper.net/problemreport/PR1824750