External gateway protocol RGP

The external gateway protocol RGP 1. Introduction The external Gateway Protocol BGP is called the Border Gateway Protocol. Why does the external gateway not use the internal gateway protocol? The main difference is that BGP uses different environments. There are two main reasons: first, the large size of the Internet makes routing between AS very difficult. Think about it. It is obviously unrealistic to Use OSPF to create a very large database. 2. Related policies must be taken into account for routing between. For example, a security problem or a route on the path does not allow passing through a non-AS datagram. Therefore, BGP can only be used to find a route that can reach the target network (not in circles), rather than finding the best route. BGP adopts the path vector routing protocol, which is different from distance vector protocol and link status protocol. 2. When BGP is configured, the administrator of each autonomous system must select at least one router as the "BGP spokesman" for the autonomous system ". To exchange route information, a BGP spokesman must establish a TCP connection and then exchange BGP packets to establish a BGP session and use BGP to exchange route information. For example, in addition to the BGP protocol, each BGP spokesman must run the internal gateway protocol used by the autonomous system, such as OSPF or RIP. The Network accessibility information exchanged by BGP is a series of autonomous systems that reach a certain network. After the BGP spokespersons exchange network accessibility information, the BGP spokespersons find a better route to the autonomous systems from the received route information. The connectivity diagram of the autonomous system constructed by the BGP spokesman is a tree structure with no loops. For example, an example of a BGP spokesman switching the path vector is provided. The BGP spokesman for AS2 of the autonomous system informed the BGP spokesman for the backbone network: "To reach the network N1, N2, N3 and N4, you can pass through as2 ." After receiving this notification, the backbone network sends a notification: "To reach the N1, N2, N3, and N4 networks, you can follow the path (AS1, AS2 )." Similarly, the backbone network also sends a notification: "To reach the network N5, N6, and N7, you can follow the path (AS1, AS3 )." The path vector information is used here, so it can effectively avoid the phenomenon of circles. For example, if a BGP spokesman receives a path notification from another BGP spokesman, it must check whether the autonomous system is in this path. This path cannot be used in this path. It can be seen that the number of nodes in the BGP exchange route information is an order of magnitude of the number of autonomous systems, which is much less than the number of networks in the autonomous system. At the same time, finding the correct path is to find the correct BGP spokesman. 3. When BGP packets are running, the BGP site switches the entire route table. However, you only need to update the changed part in the future. This reduces the consumption of network bandwidth. BGP has four message formats: ① OPEN (OPEN) packets-used to establish a relationship with another adjacent BGP speaker, which is the communication initialization. ② UPDATE message: used to advertise information about a route and list multiple routes to be revoked. ③ KEEPALIVE (keep-alive) packets are used to periodically verify the connectivity of neighboring sites (compared with TCP keep-alive ). ④ NOTIFICATION message: used to send detected errors. If a BGP spokesman wants to periodically exchange information with another BGP spokesman, you must first establish a relationship because the other BGP spokesman may have heavy load and therefore do not want to establish a relationship. When the relationship is established, an OPEN packet is sent, and another BGP replies to the KEEPALIVE packet. Once the relationship is established, maintenance is performed, and KEEPALIVE packets are sent. The UPDATE message can cancel the previously notified path or add a new path. 4. Comparison of RIP, OSPF, and BGP uses UDP, OSPF uses IP, and BRP uses TCP. What are the advantages of doing so? Why? Why does RIP periodically exchange information with neighboring sites while BGP does not? RIP only exchanges information with neighboring sites. Although UDP does not guarantee reliable delivery, it has low UDP overhead and can meet the requirements of RIP. Due to the use of UDP, RIP periodically exchanges information with neighboring sites. To overcome the disadvantages of unreliable UDP. OSPF uses reliable flood methods. Therefore, IP addresses are used directly. The advantage is that the flexibility is good and the overhead is low. RIP needs to exchange the entire route table and update information. Therefore, to ensure correctness, TCP is used. Because RIP uses TCP, reliable delivery can be ensured and periodic interaction information is not required.