Dynamic Routing Protocol (RIP)

Dynamic Routing Protocol (RIP) 1. introduction for Static Routing, route table items are generated by default when the interface is configured (for directly connected interfaces ), and add table items (usually from the system self-pilot program file) through the r o u t e command ), or generate table items through I C M P redirection (usually when an error occurs by default ). This method is feasible when the network is small and there is only one connection point with other networks and there are no redundant routes. If none of the above three conditions are met, dynamic routing is usually used. Dynamic Routing mainly uses r I p (Routing Infromation Protocol). Most t c p/I P implementations provide this widely used Protocol. 2. Dynamic Routing when the adjacent routers communicate with each other to inform the other router of the network currently connected, a dynamic routing occurs. Vrouters must communicate with each other using the routing protocol. There are many such routing protocols. A process on a router is called a routing daemon. It runs the routing protocol and communicates with neighboring routers. 3. RIP: The RIP packet is contained in the UDP datagram. Here we will only summarize version 1. for version 2, refer to TCP/IP (v1 ).
The command field is 1, indicating a request, and 2, indicating a response. There are two additional discard commands (3 and 4), two informal commands: Round Robin (5) and round robin (6 ). The request requires other systems to send all or part of the route table. The response contains all or part of the route table of the sender. The version field is usually 1, and the r I P of version 2nd sets this field to 2. The address family (2 for the I p address, the value is 2), the I P address, and the corresponding measurement. Later in this section, we can see that the r I P measurement is based on the Skip count. Up to two or five routes can be advertised for r I P messages in the format of 2 0 bytes. The upper limit of 2 5 is used to ensure the total length of the r I P packet is 2 0 × 25 + 4 = 504, less than 5 1 2 bytes. Because each packet can carry up to two or five routes, multiple packets are often required to send the entire route table. Note that the maximum hop count is 1 5, which means that r I P can only be used in A S (Autonomous System) where the maximum hop count between hosts is 1 5. The measurement is 1 6, indicating that no route arrives at the I P address. The maximum number of hops is 15, which limits the size of the RIP network. 3.2 running steps Let's take a look at the results of the normal operation of the R o u t e d program using the r I P protocol. The u d p port number commonly used by r I P is 5 2 0. 1) initialization: when starting a routing daemon, it first determines which interfaces are started and sends a request message on each interface, requiring other routers to send the complete route table. In point-to-point links, this request is sent to other endpoints. If the network supports broadcast, such requests are sent in the form of broadcast. The command field of this request message is 1, but the address series field is set to 0, and the measurement field is set to 1 6. This is a special request message that requires the complete route table at the other end. 2) receive the request. If the request is a special request just mentioned, the router sends the complete route table to the requester. Otherwise, each table item in the request is processed: If a route is connected to the specified address, the measurement is set to our value, otherwise, set the measurement to 1 6 (measurement 1 6 is a special value called "infinity", which means the route does not reach the destination ). Then send back the response. 3) receive the response. If the response takes effect, the route table may be updated. New table items may be added, existing table items may be modified, or existing table items may be deleted. 4) regular route-based updates. Every 3 0 seconds, all or some routers send their complete route tables to adjacent routers. The sending route table can be in the form of broadcast (such as over Ethernet) or sent to other endpoints of the point-to-point link. 5) trigger the update. Update a route whenever its measurement changes. Instead of sending the complete route table, you only need to send the changed table items. Each route has a related timer. If the system running r I P finds that a route is not updated within 3 minutes, set the measurement of the route to infinite (16) and mark it as deleted. This means that you have not received any updates from the router that advertised the route within 6 3 0 seconds. After 6 0 seconds, the route will be deleted from the local route table to ensure that the failure of the route has been spread.