Overview of Common Internet Router protocols

Many people may be familiar with the term Router Protocol, especially the rapid development of the network, the Internet Router Protocol is constantly improved, and many new features are also emerging. The wide application of information technology in various fields has promoted the rapid development of information exchange networks. Internet is the biggest beneficiary.

The main node and device of the Internet network are routers. The router technology determines data forwarding through routing. The forwarding policy is called routing selection routing), which is also the origin of the router name router, the forwarder ). The forwarding method can be manually specified, but the workload is too heavy to adopt flexible policies. As a result, the dynamic Router Protocol came into being, through propagation, analysis, calculation, and selection of routes, routing discovery, routing selection, Routing Switching, and load balancing.

RIP, OSPF, and BGP protocols

A large number of router protocols running on the Internet are RIP, OSPF, and BGP. RIP and OSPF are internal gateway protocols. They are applicable to the operation of the unified Router Protocol of a single ISP. A network operated by an ISP is called an autonomous system ). BGP is a router protocol between autonomous systems and an external gateway protocol.

RIP is the most time-consuming Router Protocol and the simplest Router Protocol. It is short for "Routing Information Protocol" and mainly transmits route information route tables) to broadcast routes: Route tables are broadcast every 30 seconds to maintain the relationship between neighboring routers, calculate the route table according to the received route table. RIP is easy to run and applies to small networks. RIP is still used on the Internet.

OSPF protocol is short for "Open Shortest Path First. "Openness" is aimed at the "private" Router Protocol of some manufacturers at that time. It is precisely because of the openness of the Protocol that OSPF is powerful and widely used today. It obtains network information by transmitting link state connection information, maintains a directed network topology, and uses the Minimum Spanning Tree Algorithm SPF to obtain the route table. OSPF is a relatively complex Router Protocol.

In general, OSPF and RIP are vro protocols in autonomous systems and are suitable for a single ISP autonomous system. Generally, the entire Internet is not suitable for running a single Router Protocol, because each ISP has its own interests and is unwilling to provide detailed routing information of its own network. To ensure the interests of ISPs, the Standardization Organization has developed the BGP protocol between ISPs.

BGP is short for the "Border Gateway Protocol". It processes route transmission between ISPs. It is characterized by a variety of routing policies, which cannot be achieved by RIP, OSPF, and other protocols, because they require global information to calculate the route table. BGP adds a certain policy to the router at the ISP border, selects a filter route, and sends routes such as RIP, OSPF, and BGP to the peer. The global and extensive Internet is an instance where BGP processes routes between multiple ISPs. The emergence of BGP has led to a major revolution in the Internet. It connects multiple ISPs and truly becomes a global network. The side effect is the Internet route explosion. There are about 60000 routes on the Internet, which is still the number after "aggregation. To configure BGP, you need to have a good understanding of user requirements, network conditions, and BGP protocols. Also, you need to be very careful that BGP runs at the core. Once an error occurs, the loss may be huge!

To adapt to One-to-multiple multi-point transmission applications in the Internet, such as weather forecasts and network conferences, a new transmission mode, multicast, has emerged ). Multicast is suitable for one-to-many transmission environments and can also be applied to multiple-to-one transmission environments. Multicast Forwarding is mainly determined by the vro. The vro determines the downstream of the so-called downstream: determines whether there are host users) through the MulticastclientIGMP protocol, and determines whether there are indirect users, that is, through the "downstream" vro band members, the downstream vro sends messages through multicast vro protocol, and the vro determines whether to forward data to the downstream. We can see that the application of the multicast Router Protocol in the second method is the key to large-scale Network Multicast forwarding. The multicast Router Protocol should at least correctly notify the team members and form a globally unified routing topology.

The dense mode is suitable for small networks. It is assumed that there are very "intensive" members across the network, and the working strategy of broadcast + pruning is adopted. It is assumed by default that data is forwarded to all downstream instances. This interface is removed from the downstream list only after a specific pruning message is received from a downstream instance. Generally, the forwarding path should be a tree with "Source" as the root and group members as the branches and leaves. Vro protocols in dense mode include DVMRP, MOSPF, and PIMDM.

Sparse Mode is widely used on the Internet. After all, for the Internet, no more than 1% of machines need to receive any multicast application. In Sparse Mode, by default, all machines do not need to receive multicast packets. Only those who explicitly specify the need are forwarded. This is indeed applicable to the "sparse" consideration. Currently, the main idea of forwarding all Sparse Mode protocols is that all similar packets are forwarded in the same path, that is, they are first sent to an aggregation point or a core ), then forward the data along the shared tree with the group members taking the aggregation point as the root. Sparse router protocols include PIMSM and CBT.

The router protocol supports IP addresses and the Internet. Without the router protocol, the Internet will be a chaotic world, and it cannot be as convenient and fast as it is today. Multicast has created a new development prospect and will become the main force in guiding the future of the Internet.