Routing protocols

Routing protocols are designed to collect data on network topologies. The main task of routing, or, in other words, the selection of the most optimal route, is usually solved by analyzing special tables that are located in all end network nodes and routers.

In order to automatically build the required tables, the routers change with each other certain information using specially designed service protocols, they are called “routing protocols”. These include the NLSP, RIP, OSPF protocols, which must be distinguished from the network, for example, such as IP.

Using appropriate protocols, routers gradually map a network connection. Already on the basis of this data, for each of the network numbers, a decision is made which particular router needs to transmit packets that are sent to this network so that the route ultimately turns out to be the most rational. The results of decisions made are recorded in the routing table. When network configurations change, some entries automatically become invalid. Then packets that are sent on false routes may be lost or looped. The quality of the work directly depends on how quickly the routing protocols bring the contents of the table to the real situation on the network.

Routing protocols have several classifications. They can be either single-step or multi-step, static, dynamic, class, classless. In addition, such protocols can be external and internal. In one-step when choosing the most rational route, only the nearest router is initially determined, and not their entire sequence. Static routes are entered manually by the network administrator . Usually they are used in small networks that have a simple and understandable structure. The plus, of course, is the ease of configuration, the absence of any traffic loss for the transmission of routing information and low resource requirements. But if there are changes in the network configuration, you have to change the routing table on all hosts manually. However, dynamic routing protocols are very popular.

BGP is one of the main routing protocols on the Internet. It is intended for the exchange of route data between large autonomous systems, therefore, in addition to standard information, it transfers route data to autonomous systems. BGP chooses the best route based on the rules that are accepted on the network and does not use technical metrics in its work. This protocol also uses route summarization to reduce routing tables. Currently, the fourth version of the protocol is valid.

OSPF is also a fairly popular dynamic protocol. It is based on technology that tracks the status of the channel and uses the Dijkstra algorithm for its work . Its advantages include a high convergence rate, the most efficient use of bandwidth and support for variable-length network masks .

The RIP protocol is one of the oldest, which, however, is quite widespread to this day. It is used in small networks with a simple structure. The protocol is easy to operate and install. His work is based on distance vector algorithms. When using RIP, all entries in the routing tables contain either the network address or the destination host.

Routing protocols are quite diverse, there are a large number of them, and each has its own advantages and disadvantages.

Source: https://habr.com/ru/post/C1391/


All Articles