On the virtual routing cluster technology of gigabit switching routers

First, the introduction:

Currently, IP has become a routing protocol for most backbone network products. In some network environment, the user's request to the network is very high, any shutdown and storage and transportation loss will cause serious impact to the user. For example:

1. Internet service providers provide web host devices, and users must be guaranteed 99.9999% uptime in order to make their Web servers available to the public.

2. The Process Control application must be able to access its controlled system in a timely manner, otherwise it may result in serious loss of control;

3. Sometimes, an application running on an IP host times out, which can have a bad effect if the business operation requires a higher level of network application.

More and more IP hosts use DHCP to specify their IP addresses. However, many IP hosts use manual configuration as the only means of discovering gateway routing addresses. Some hosts use the Gateway detective method to obtain, but generally do not promote the use of this method (RFC1122), dynamic Ping Gateway is also prohibited to use. The ICMP Routing Discovery protocol allows routers to be discovered through IP hosts, but is not yet widely used.

This means that most hosts cannot quickly know whether the router and its associated LAN connection have failed, and that it takes a long time for the IP host to detect the link failure to swap with the alternate router.

Therefore, it is essential for any device to improve the effectiveness of the network. Using virtual routing cluster technology, some problems can be solved effectively. This paper introduces the SSR Gigabit switched routing cluster Technology of Cabletron company, and makes some discussion.

Second, how to do?

Using Virtual routing clustering technology, the Cabletron multiple SSR routers located in the same LAN can be defined as the cluster of routing backups between each other, and the SSR routers in the cluster use the IETF Virtual Redundancy routing protocol to realize the mirroring between the routing switches and the LAN of each router. If a router or the LAN connection to which it is connected fails, the other routers automatically bypass the failed router to reroute at the point of failure, and the route recovery time is within seconds, so this is a transparent application running on the IP host.

(Figure I)

1, the concept of VRRP

As shown in figure, VRRP can provide a backup router when the workstation's default gateway fails, VRRP can create a virtual router with a virtual MAC address and virtual IP address.

Admittedly, we can add multiple default gateways in Windows98, but this feature can only be used when the host is started. In other words, the first default gateway in the Gateway list at host startup uses ARP for address resolution, and if parsing fails, address resolution is continued for other gateways in the list.

When the host finds the MAC address of the default gateway, the gateway suddenly becomes paralyzed. Only the new host can be restarted to continue looking for the next default gateway.

Using VRRP can solve this problem because it is dynamic and if the primary router is paralyzed, the backup router automatically replaces the primary router. The host is unaware of the slightest difference.

Virtual routing cluster technology is integrated into SSR software as a module to ensure the practicability of network application, and improve the reliability and adaptability of the network with higher application requirements.

The Cabletron Virtual routing clustering technology follows the standard protocol of VRRP (Virtual Redundancy routing protocol) developed by the IETF (Internet Engineering Task Force).

(Figure II)

In figure II, IP host H1, H2, etc. are connected to a LAN with multiple routers (R1, R2, etc.), and R1 and R2 provide connections to the destination D1. Normally, the host H1, H2, and so on, only configure a single router R1 or R2 routing ip address. The problem is that when any one of the routers fails, one or more hosts H1, H2, and so on, drive the connection with the destination D1. For general routing protocols such as RIP, OSPF, and so on, host H1, H2 and so on find that routing failures take a long time (about 40-90 seconds) and even cause TCP connections to be interrupted. When the VRRP protocol is used in virtual routing clustering, the routing failure recovery in the network takes only a few seconds, which greatly enhances the effectiveness of the network and provides routing redundancy for the critical backbone of the network application.