2 sites, conn via T1, internet connetion on both sides - REDUNDANT POSS.?

[jkeegan123]

If there are 2 sites that each have their own internet connection, with a ppp link in the middle (cisco 1720 routers), can the internet connectivity be made to fail over to the other site automatically?

Given the following scenario:

Site 1:Cisco 1720 router

--------------------------------------

FastEthernet0: 10.0.1.1 255.255.255.0

Serial0: 172.16.1.1 255.255.255.252

encap HDLC

ip subnet-zero

Router EIGRP 100

network 172.16.0.0

network 10.0.0.0

no auto-summary

ip route 0.0.0.0 0.0.0.0 10.0.1.254 (sonicwall firewall)

client configuration: DHCP

10.0.1.50-10.0.1.200

SNM: 255.255.255.0

DG: 10.0.1.1

Site 2:Cisco 1720 router

--------------------------------------

FastEthernet0: 10.0.2.1 255.255.255.0

Serial0: 172.16.1.2 255.255.255.252

encap HDLC

ip subnet-zero

Router EIGRP 100

network 172.16.0.0

network 10.0.0.0

no auto-summary

ip route 0.0.0.0 0.0.0.0 10.0.2.254 (sonicwall firewall)

client configuration: DHCP

10.0.2.50-10.0.2.200

SNM: 255.255.255.0

DG: 10.0.2.1

How can these sites internet connections be made redundant? For

example, sine the routers on each side each have a default route to

the local firewall (a sonicwall device), how can I configure these

routers so that, if the sonicwall locally is unresponsive (i.e. the

local internet connectiion is down), the router automatically routes

the connection to the default gateway of the router on the other side

of the WAN?

There has to be an easy way to make this happen. Any help is

appreciated.

Thanks!

Joe Keegan

jkeegan@constructuretech.com

[thiland]

The problem I see is that the internet connection is not directly connected to your Cisco 1720, and so your router doesn't know if the internet connection goes down.

Is there a way to run a routing protocol between your Sonicwall and Cisco router? If you could run, say RIP between them, then your Cisco router would know when the internet connection is down. Then redistribute RIP into EIGRP, and both of your WAN routers could have a feasible successor to 0.0.0.0.

It will be interesting to see other solutions posted.

[Richard Burts]

An easy solution to this kind of issue is to use floating static routes. A floating static route is a static route with an administrative distance higher than normal. So long as the normal static route is usable it is used but if it becomes unusable then the floating static route is used. In your case your site 2 1720 could have configured ip route 0.0.0.0 0.0.0.0 172.16.1.1 250. The 250 at the end is what makes it a floating static route.

The difficulty with this solution is what would make the original static route (ip route 0.0.0.0 0.0.0.0 10.0.2.254) unusable. If your router Fast Ethernet interface went down then the floating static could be used. But the chances are much greater that the difficulty will be at the sonicwall than that your interface would go down. If you could set up some kind of dynamic routing so that the sonciwall advertised a default route to you, then if there were a problem with the firewall and the default stopped being advertised then the floating static route could work.

HTH

Rick

[vcjones]

This is actually fairly easy to do, but it is not straightforward. You need to use SAA RTR to detect when one of the SonicWalls is no longer useful, and have that knowledge tracked in a static route defining your default route.

Do not use a floating static route to the other end of the router-to-router link for your backup static route. (Consider what happens if both SonicWall routers go down). Instead, redistribute the tracked static route into your routing protocol so the other router can learn that you have a default route, and set the admin distance on the tracked static route to more desirable than the routing protocol.

Warning: the SAA RTR capabilities required are only available today in 12.3T, which puts you on the bleeding edge of IOS releases, so budget extra time to work around any bugs you encounter.

Good luck and have fun!

Vincent C Jones

www.networkingunlimited.com

[jkeegan123]

I started getting into the SAA RTR commands. I am currently trying to run an SAA RTR ping and track that ping for reachability, but my IOS only knows how to track interfaces or routes, not RTR objects.

Any thoughts as to why? I'm running 12.3(2)T7 on a 1700 router.

Here's what I did with RTR:

rtr 1

type echo protocol ipIcmpEcho 167.206.112.3 (internet DNS server)

rtr schedule 1 life forever start-time now

then, as per someone's advice, I tried to do:

track 123 rtr 1 reachability

but my track command won't accept that.

Am I missing anything?

[vcjones]

This feature was introduced in one to the 12.3T releases, you may not have a recent enough version of 12.3T (current at 12.3(8)T, IIRC).

Vincent C Jones

www.networkingunlimited.com