DELL 8132 ユーザーズマニュアル

Virtual Router Redundancy Protocol Commands

An end station running IP needs to know the address of its first hop router. 

While some network administrators choose to install dynamic router 

discovery protocols such as DHCP, others prefer to statically allocate router 

addresses. If the router identified by such a statically allocated address goes 

down, the end station loses connectivity. The Virtual Router Redundancy 

Protocol (VRRP) is designed to provide backup for the failing router without 

requiring any action on the part of the end station. It is based on the concept 

of having more than one router recognize the same IP address. One of the 

routers is elected the master router and handles all traffic sent to the specified 

virtual router IP address. If the master router fails, one of the backup routers 

is elected in its place and starts handling traffic sent to the address. This 

change is transparent to end stations.
VRRP increases the availability of the default path without requiring 

configuration of dynamic routing or router discovery protocols on every end 

station.
Multiple virtual routers can be defined on a single router interface.

RFC 3768 specifies that a router may only accept IP packets sent to the 

virtual router’s IP address if the router is the address owner. In practice, this 

restriction makes it more difficult to troubleshoot network connectivity 

problems. When a host cannot communicate, it is common to ping (send an 

ICMP Echo Request) the host’s default gateway to determine whether the 

problem is in the first hop of the path to the destination. When the default 

gateway is a virtual router that does not respond to pings, the operator cannot 

use this troubleshooting technique. Because of this, it has been common for 

VRRP implementations to respond to pings, in spite of the prohibition in the 

RFC. The IETF has recognized the issue, and a draft revision of the VRRP 

2CSPC4.X8100-SWUM102.book  Page 1379  Friday, March 15, 2013  8:56 AM