Introduction and classification of Dynamic Routing Protocols

Introduction and classification of dynamic routing protocols different routing protocols use different measurements, and sometimes multiple measurement hops bandwidth Load Measurement reflects the traffic usage of links along the way; unlike the number of hops and bandwidth, the load on the path changes, so the measurement also changes. If the measurement changes too frequently, the routing swing (the Optimal Path changes frequently) may occur frequently; the routing swing will negatively affect the CPU, bandwidth of the data link, and network stability of the router. The latency reliability measurement is used to measure the possibility of link failure under certain circumstances; reliability is changed or fixed; the number of link failures or the number of errors received within a specific interval is an example of variable reliability measurement; the fixed reliability measurement is based on the known amount of a link determined by the Administrator. The cost is a general term used to describe the advantages and disadvantages of a route. The minimum cost (maximum cost) or the shortest (longest) only refers to a view of the routing protocol on the path based on its specific measurement. The administrator can manually define the cost to converge dynamic routing. The selection protocol must contain a series of processes, these processes are used by the router to notify other routers of the local direct connection network and receive and process Similar information of the vro, which is the information that the relay receives from other vrouters. In addition, the routing selection Protocol also needs to define the measurement that determines the optimal path. For the routing protocol, another criterion is that the reachable information in the route tables of all routers on the interconnected network must be consistent. The process of making all route tables consistent is called "convergence "; the sum of the time it takes for information sharing across the network and for all routers to calculate the optimal path is the "convergence time" Dynamic Routing Protocol Classification 1 A group of internal and external gateway protocols under the common management domain A set of routers running the same routing selection protocol is an Autonomous System (Autonomous System, AS); in the Internet, an autonomous system is a unit (an enterprise, city, or isp) that has the right to determine which Routing Protocol the system uses. With the development of the network, the above definition of AS is no longer very accurate. The development of the network leads to frequent network convergence between networks, resulting in more and more routing protocols used in the same autonomous system, therefore, autonomous systems should be defined as interconnected networks under the joint management; Internal Gateway Routing Protocol (IGP): used for decision-making routing in a single autonomous system; Package Including RIP, OSPF, IS-IS, and other external Gateway Routing Protocols (EGP): used to execute routes between multiple autonomous systems. For example, bgp igp is used to solve internal communication of, EGP is the origin of the 2-distance vector and link-State Routing Protocol distance vector name for inter-AS communication because the route is a vector (distance, direction) the distance is defined according to the measurement (using 'hops 'as the measurement), and the direction is defined according to the next hop router; for example, the "downward hop in the direction of router X, based on which the distance is 5 hops," indicates that each router learns the route information they observe from the neighboring router, then, we will advertise the route information we observed. Because each router depends on the neighboring router in the information, and the neighboring router learns the route from their neighboring router, therefore, distance vector routing is sometimes referred to as "Routing Based on rumors". The information used by the distance vector routing protocol can be compared to the information provided by the 'road; the link status routing protocol is like a 'road path'; the link status router is not prone to spoofing and makes wrong routing decisions, because it has A complete network diagram. The link status is different from the distance vector's method of routing selection based on rumors because the link status router obtains first-hand information from the peer router; each vro generates information about itself, the local smart network, and the status of these links. The information is transmitted from one vro to another, and each vro backs up the information, however, no information will be changed. The ultimate goal is that each router has the same information about the interconnected network, and each router can independently calculate its optimal path; to understand the concept of several distance vector routing protocols, regular update means literal update. It should be noted that if the update information is sent too frequently, blocking may occur, however, if the update information is not sent frequently, the network convergence time may be long and cannot be accepted by the neighbor distance vector routing protocol to send the update information to the adjacent router, the neighbor is dependent on passing update information to its neighbor. Therefore, the distance vector routing protocol is said to be "update by Hop". When the router is activated on the network for the first time, how do vrouters find other vrouters? How does it declare its own existence? The simplest method is to send an update message to the broadcast address (255.255.255.255.255 in an IP address). A neighbor router using the same route selection protocol will receive the broadcast packet and take corresponding actions; hosts and other devices that do not care about route update information will discard this packet. The full route table updates most distance vector routing protocols use a very simple way to tell neighbors what they know, this method broadcasts its entire route table. After receiving the update information, the neighbor collects the information required by the host, and the other information is discarded.