CCNA - Static route: Local interface vs Next hop

Author

Static route: Local interface vs Next hop

Warfare

2004-08-26, 11:49 pm

I have an unanswered question about adding static routes. I have argued about this with few fellow students but we couldn't come to a conclusion.

When adding a staic route:

#ip route 0.0.0.0 0.0.0.0 ......

Why would you use next hop (gateway) address instead of local interface or vice versa?

I understand that by using locl interface serial 0/0 the admin distance will be 0 (directly connected) and 1 if the next hop (gateway) address has been used.

My argument was that adding the route using the local interface will send the packets on that serial link regardless of anything.

While adding the route using the gateway, ensures that the next hop is reachable first then adds it to the routing table, and should it become unreachable someday, the route will be marked as inaccessible.

Does that make sense?

stnosc

2004-08-27, 9:08 am

quote:
Originally posted by Warfare
Why would you use next hop (gateway) address instead of local interface or vice versa?

Here's the scenerio:

You have a perimeter router with two paths out to the WAN. One path you will use for a backup link, the other will be active. You are not using a routing protocol, but default static routes. The way to do this is by adjusting the administrative distance on the default route for the backup higher than the one for the active link. If the active link goes down, the backup link takes over.

If you are using exit interface in your default routes (i.e 0.0.0.0 0.0.0.0 s0/0),
the link will not swap to the backup unless the interface goes down (Serial0/0 is down, line protocol is down). The problem is that if your distant node is having problems, your interface could still be up even if your line protocol is down (Serial0/0 is up, line protocol is down). As long as your physical interface is up, the route remains active even though you are unable to pass traffic to the distant node.

If you use the next hop (your distant node) in your active default route (i.e 0.0.0.0 0.0.0.0 10.10.10.18), when the distant node is having problems, your router will know that it can't pass traffic to the node because the gateway of last resort will not be reachable. Once that happens, your standby default route (the one with the higher administrative distance) will kick in, your gateway of last resort will change to the next hop of your backup link, and traffic will flow.

Hope this wasn't to long and drawn out....

smrkdown

2004-08-27, 10:53 am

Are you only talking about default routes, or all static routes?

stnosc

2004-08-27, 1:00 pm

The example he gave was a default route, so I assumed that's what he was talking about.

Warfare

2004-08-27, 7:55 pm

Yes I was talking about the default route specifically using (ip route 0.0.0.0 0.0.0.0 xxxx)

I think I see some logic in your answer. I was thinking something along those lines but wasn't sure of it or how to explain it exactly.

Thanks

Just Visiting

2004-09-01, 5:28 am

Putting your exact question aside and looking at it more generally, investigate what happens when you make the static route next hop an ethernet interface instead of a serial.

-JV

Warfare

2004-09-01, 9:51 am

Now I think of it, I already have my default route on Router1 defined as next hop to an ADSL/router/hub gateway addess 192.168.0.1 using ethernet connection, RIPv2 used on both 2500 routers and the ADSL Hub.

The static route is propagated and the update is sent to Router2. If I try to duplicate the same scenario on the academy lab, this time using a loopback0 interface, the static route won't propagate even if the command default-information originate is being used.

I will test the local interface vs next hop scenario when the link is up and down.

Thanks for the tip

Sponsored Links