Dynamic routing

小乔

It is obtained by exchanging information among interconnected routers and then optimizing it according to certain algorithm

1、 Routing basis
Default management distance
Direct interface  0
Static routing  1
EIGRP summary Routing  5
EBGP  20
EIGRP   90
OSPF  110
ISIS  115
RIP  120
ExEIGRP  170
IBGP  200
Unknown  255

1. Administrative distance_ Distance (AD) measures the credibility of routing information from neighboring routers. The management distance is an integer value from 0 to 255, 0 is the most reliable, and 255 means that no traffic will pass through this route.
2. Metric, a method to determine the best path from a routing protocol to the destination network. When a router has multiple paths to reach a certain destination network, the routing protocol must determine which one is the best and put it into the routing table. The routing protocol will calculate a number for each path, which is the measurement value. Usually, this value has no unit. The smaller the measure, the better the path. However, different routing protocols have different ways to define metrics, so different routing protocols may choose different optimal distances.
Note: ① if a router receives two updates to the same remote network, the first thing the router should check is ad. if one of the announced routes has a lower ad value than the other, the route with a lower ad value will be placed in the routing table.
② If two advertised routes to the same network have the same ad value, the metric of the routing protocol (such as the number of hops or the bandwidth value of the link) will be used as the basis for finding the best path to the remote network. Routes with lower metrics that are advertised will be submitted to the IP routing table.
③ If two advertised routes have the same ad value and the same metrics, the routing protocol will use load balancing for the remote network (that is, the packets it sends will be equally distributed to each link).

2、 Characteristics of distance vector routing algorithm
1. Only exchange information with neighboring routers;
2. The information exchanged by the router is all the current routing information of the router, namely its own routing table (a fundamental difference between the distance vector routing protocol and the link state routing protocol is that the distance vector routing protocol exchanges routing information, while the link state routing protocol exchanges link state information);
3. Exchange routing information at fixed intervals (for example, Rip is 30 seconds, which is actually a floating value).

3、 The limitation of routing loop
Due to the problem of routing algorithm or various abnormal factors in the network, the message sent from a router in the network returns to the router after a period of time. Such a process is called routing loop.
Therefore, the maximum number of hops that can reach 16 is considered as the maximum number of hops that can not be reached by any rip protocol.
2. Horizontal segmentation. Routing protocol distinguishes which interface gets the network routing information. Once this judgment is confirmed, it will not send the information about this routing from this interface again.
3. Route poisoning, defining the maximum value solves the problem of route loop to a certain extent, but it is not thorough. It can be seen that the routing loop still exists before the maximum value is reached. Therefore, route poisoning can completely solve this problem. The principle is as follows: suppose that there are three routers a, B and C (the topology is that router C is directly connected to network x, B is connected to C, and a is connected to b). When network x fails and cannot be accessed, router C will send relevant route update information to neighbor router B, and mark the metric value arriving at network x as initify or unreachable (sometimes regarded as infinity) table entry to trigger A route poisoning tells neighbor router B that network x is out of reach. After receiving the poisoning message, router B marks the link routing table entry as infinite, indicating that the path has failed. It also notifies neighbor router a, poisons each router in turn, and tells neighbor network X that the network has failed, so as to avoid the probability of routing loops.
4. Poison Reversion: in combination with the above example, when router B sees that the metric of arriving at network x is infinite, it sends an update information called poison reversion to router C, indicating that the network x is unreachable. This is a special case beyond horizontal segmentation, so as to ensure that all routers receive the poison routing information.
5. Keep closed, which can prevent regular update messages to restore a constantly open and close (called flipping) route, that is, to set a keep invalid timer.
6. Trigger update: the router that detects the network change will immediately send an update information to the neighbor router, and in turn generate trigger update notification to their neighbor router, so that the router on the whole network can receive the update information in the shortest time, so as to quickly understand the changes of the whole network.
Summary of Cisco knowledge points