Which address type is commonly used in the BGP neighbor command for iBGP sessions?

• A. The MAC address of the neighbor's interface
• B. The router ID
• C. A loopback address is generally used in the BGP neighbor command
• D. A multicast address

Answer: (C)

In iBGP peering, using a loopback address for the neighbor is the common practice because it gives you a stable, always-up endpoint for the BGP TCP session. Loopbacks are logical interfaces that remain up regardless of the state of any physical link, so the peering plane isn’t as prone to flapping if a router interface goes down or is reconfigured. To make this work, you typically ensure reachability to the neighbor’s loopback (via IGP or static routes) and configure the session to source from the loopback on both sides (often using update-source loopback). This combination keeps the session reliable and predictable.

Other address types don’t fit as well for this purpose. A MAC address is a layer 2 construct and isn’t used to establish BGP sessions, which run over IP/TCP. The router ID is a unique identifier for BGP itself, not the address used to form a neighbor relationship. A multicast address isn’t appropriate for unicast BGP peering, which relies on a specific unicast IP path.