cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
368
Views
0
Helpful
3
Replies

Router redundancy

christianIP
Level 1
Level 1

Hi,

In a Cisco 2610 router with three faEthernet interfaces (one for the LAN and two for internet access with two different providers), how could redundancy be enabled? I am aware HSRP can be used but it requires two or more routers, also VRRP can be used but this also requires two or more routers. How can redundancy be provided in this scenario? The idea is to have one faEthernet interface with one ISP and the other one with a different one, when one of them goes down, fault redundancy takes places transparently and automatically, when both faEth interfaces are up and running, all clients will be using both so service will be faster.

Your input will be highly appreciated.

Thanks.

3 Replies 3

spremkumar
Level 9
Level 9

hi

There are few concerns on the kinda requirement you have in place.

you need to have your own public ip space to do the load balancing on both the links.

also you need to have BGP running between your SP and your router to achieve this.

Since you cant route the ip space of one SP on the other which my result in loop i dont think its possible if you are using ip space from any of the SP.

But in wrost case you can make use of the links for redundancy purpose that is when one of the link goes down you can make use of the other link..

regds

Fernando_Meza
Level 7
Level 7

Hi dude ..

You can definetely have redundancy with that setup using BGP but you can't have load balancing between the links as BGP only supports one best path.

You need to establish 2 neighbor relations to two private AS ( one per ISP ) and ask them to inject a default route in to it ONLY. In that way BGP will do the redundancy part when one of the links is down.

The link below provide an example of what you are looking for

http://www.cisco.com/warp/public/459/40.html#conf4

I hope it helps !!!

Hi Fernando

FYI with BGP weight attribute conjunction with policy routing and AS Path prepends we did achieve both outgoing and incoming load balancing based on the networks.

regds