Comprehensive and Detailed 250 to 350 words of Explanation From VMware Cloud Foundation (VCF) documents:

InVMware NSXnetworking, the Tier-0 Gateway'sRouting Table(RIB) is the definitive source for determining how to reach BGP neighbors. A common point of confusion occurs when an administrator can "ping" a neighbor but the BGP state remainsIdleorConnectwith a "No route to peer" error.

This symptom specifically points to the"Scope"setting of a static route. In NSX, when a static route (such as the default route 0.0.0.0/0) is created, the administrator can define theScopeto be a specific uplink segment or interface. If the scope is set exclusively to theVLAN 100segment, the Tier-0 Gateway will only install that route into the forwarding table for the Service Router (SR) component associated with the VLAN 100 interface.

Because the default route is the only path the Tier-0 has to reach non-local networks (or even other local subnets not directly attached), the BGP process for the neighbor at192.168.101.1(VLAN 101) checks the routing table for a path. Since the only available route is scoped strictly to VLAN 100, the Tier-0 determines it has "No route" to reach the neighbor in VLAN 101. BGP requires a valid entry in the routing table for the neighbor's IP before it will even attempt to initiate the TCP three-way handshake on port 179.

The fact that pings succeed is due to pings often being tested from the specific interface (e.g., ping 192.168.101.1 -I fp-eth1), which bypasses the general routing table logic that the BGP control plane must follow. To resolve this, the static route scope should be expanded to include all relevant uplink segments or left as "All Uplinks," ensuring that the Tier-0 recognizes valid egress paths for neighbors on both VLAN 100 and VLAN 101.

===========