What is the practical difference between eBGP and iBGP?

The practical difference between iBGP and eBGP is that eBGP is used for exchanging routing information between different autonomous systems (AS), while iBGP is used to propagate those learned routes *within* your own AS. This fundamental distinction drives key technical differences in their operation.

eBGP peers, by default, have a Time to Live (TTL) of 1, assuming the peer is directly connected. iBGP sessions often traverse multiple internal hops, requiring the `neighbor <ip> ibgp-multihop <ttl>` command to allow the TCP session to be established. The most critical rule is that an iBGP speaker will not advertise a route learned from one iBGP peer to another iBGP peer. This loop-prevention mechanism is why you need a full mesh of iBGP sessions or a Route Reflector. eBGP, by default, modifies the BGP path attribute by prepending its own AS number to the AS_PATH, which iBGP does not do. This AS_PATH modification is fundamental to the Internet's loop prevention.

Your configuration reflects this: the external peer doesn't need `ibgp-multihop` if it's directly connected, while the internal route reflector, likely not on a directly connected network, requires it.

! eBGP Peer (directly connected, default TTL=1 is fine)

router bgp 65001

 neighbor 203.0.113.1 remote-as 65002



! iBGP Peer (multiple hops away)

router bgp 65001

 neighbor 192.168.1.1 remote-as 65001

 neighbor 192.168.1.1 ibgp-multihop 2