| Routing Protocol | Protocol No. | Port No. | Update Reliability |
| IGRP | 9 | Best effort delivery | |
| EIGRP | 88 | 1-to-1 window | |
| OSPF | 89 | 1-to-1 window | |
| RIP | UDP 520 | Best effort delivery | |
| BGP | TCP 179 | Uses TCP windowing |
IGRP, EIGRP, and OSPF are transport-layer protocols that run directly over IP. IGRP uses connectiuonless delivery for its routing updates. Routers receiving IGRPupdates do not need to acknowledge that receipt of these updates. EIGRP and OSPF have more reliability built into their update processes. They both require the acknowledgment of one update before they send another. Thus, they have a 1-to-1 window----one update and one acknowledgment.
RIP and BGP both reside at the application layer. RIP uses User Datagram Protocol (UDP) as its transport protocol; its updates are sent unreliably with best-effort delivery.
BGP uses TCP as its transport protocol. It takes advantage of the reliability mechanisms and windowing of TCP, which is important when you consider the number of routes a BGP router sends in its updates. BGP routers often carry well over 100,000 routes in their routing tables. If OSPF or EIGRP had to send updates for 100,000 routes in their 1-to-1 window, it would take a long time. Even if information for 100 routes could fit in one update, it would still take 1000 updates to send the entire table. Each update would have to be acknowledged before sending another. On the other hand, BGP routers using TCP have a 65,535-byte window limit for their updates. The routers can send information with many more routers in each update than either OSPF or EIGRP.