Introduction to Virtual Routing Redundancy protocol under Linux (VRRP)

This article describes how to implement VRRP under Linux (Virtual Router redundancy Protocol The redundancy protocol for the router). I am in the learning VRRP process of a small summary, and combined with a number of other information on VRRP some of the basic concepts, principles and how Linux to achieve VRRP an introduction.

1, VRRP Introduction

When we set up a network, in order to ensure stable and uninterrupted operation of the network, often adopt some dynamic routing protocols, such as OSPF, RIP, and so on, these routing protocols can automatically bypass a lot of network failures, such as: Router down machine. But many times we may not be able to use these high-end routing protocols:

(1), because sometimes the user side, to configure OSPF, RIP and other protocols must have upstream ISP vendor support, but many ISP vendors do not provide this service, they provide only static routes. (Sometimes the purpose of using protocols such as OSPF and RIP is to perform links and routing backups, and VRRP can completely replace their part of the work)

(2), we have more than two internet lines, they need to load balance and routing routes, routers backup (VRRP is also mainly designed for this requirement)

In order to be able to guarantee the uninterrupted and stable operation of the network, VRRP is a best choice, VRRP is the abbreviation of Virtual Router Redundancy Protocol Redundancy Protocol, which is described in detail in RFC2338.

2. VRRP concept

There are a lot of concepts in VRRP, and we'll make a one by one introduction to them with the following figure:

(1), the router RA is the virtual router Vrid 1 's main router, is responsible for the "Internal network 1" Gateway, and is the virtual router Vrid 2 backup router, when Vrid 2 's main router RB is not available, take over the work of RB.

(2), the router RB is the virtual Router Vrid 2 's main router, is responsible for "Intranet 2" gateway, and is the virtual router Vrid 1 backup router, when Vrid 1 of the main router RA is not available, take over the work of RA.

3. VRRP Terminology

(1), virtual Router: fictitious router. In our diagram, a router mirror is composed of the actual router RA and RB using the VRRP protocol.

(2), vitual Router ID: short for Vrid, is the only identity of virtual Router in the network, is a number.

(3), VRRP INSTANCE:VRRP entity. VRRP instance can be understood as a program that runs on an actual router and is responsible for some of the information required by the VRRP protocol. For example, whether this is the actual router is the main router or backup router, this router belongs to the Vrid and so on.

(4), Virtual Router IP: A virtual Router has an IP address, under normal circumstances, there is this vritual Router the main router in charge, when the main router fails by this virtual The backup router in router is in charge.

(5), Master: the main router. Whether an actual router is the main router is generally described in its own VRRP instance.

(6), Backup: Back up the router. Whether an actual router is a backup router is generally described in its own VRRP instance.

(7), Priority: priority. In VRRP instance, a priority is defined for each actual router, a number from 1 to 254 (0 and 254 are reserved in the RFC definition), and when the primary router is not available, the backup router determines who takes over the work of the primary router according to its own priority. The larger the number, the greater the weight.

(8), owner:virtual Router IP of the person in charge.

(9), authentication: validation. The actual routers in the same vrid are required to communicate, and the communication between them needs to be validated against each other. Password validation is generally used. It is also necessary to note that the actual router communication in the same Vrid uses a multicast address: 224.0.0.18.