Cisco Cisco Dynamic Fabric Automation for OpenStack Troubleshooting Guide
![Cisco](https://files.manualsbrain.com/attachments/7380d0050044647c30f5c24bbbf5d0c0b6d9bb84/common/fit/150/50/faa183d287233c52228cfea3dbc2a127fe780f60564fcb0955d9c3d1cd23/brand_logo.png)
Cisco Dynamic Fabric Automation Production Troubleshooting Guide
69
Assuming that LHR acorn3 receives the multicast traffic, shortest path tree (SPT) cutover is triggered and builds
the subsequent BGP (S,G) join.
the subsequent BGP (S,G) join.
1. LHR acorn3 will utilize a hash to determine which border leaf switch should be used as the RPF neighbor for
the (S,G) (172.16.1.101, 225.1.110.112). In this case it has chosen acorn5.
2. Both border leaf switches will receive the BGP Join, and populate fabric multicast table with the (S,G) entry,
and the selected RPF neighbor.
3. Mroute table (MRIB) will be updated on both border leaf switches with the (S,G), but only acorn5 which was
chosen as the (S,G) RPF neighbor, will become the fabric forwarder for that (S,G). The other border leaf
switch, acorn6, will be marked fabric forwarder loser. As a result, in topologies where both border leaf
switches receive the traffic, only the fabric forwarder will route the traffic into the fabric.
Detailed Troubleshooting and Verification Steps for Fabric Forwarder Election
1. In the previous section we observed that for (*, 225.1.110.112), the LHR acorn3 hash selected RPF neighbor
acorn6. Now, verify which RPF neighbor the LHR selects for the (S,G) join of (172.16.1.101, 225.1.110.112). We
notice in the example below that it will choose the other border leaf switch acorn5. The hash algorithm on the LHR
can select the same border leaf switch or a different one for the (*,G), and (S,G) joins since the hash is done on
different source addresses.
different source addresses.
acorn3%PineForest:Tree1(config)# show fabric multicast ipv4 mroute 225.1.110.112
VRF "PineForest:Tree1" Fabric mroute Database VNI: 50010
Fabric Mroute: (*, 225.1.110.112/32)
Interested Fabric Nodes: