Redundancy of two links with a dynamic routing protocol

[Tauer Drumond]

Hi,

I have two sites linked by two LAN-to-LAN links.

The first LAN-to-LAN is a 34Mbps. The second is a 12Mbps one

Im using static routes on both routers to use only the 34Mbps link, because if I load the sharing per-packet on the interfaces, some applications doesn't work properly.

And, if the 34Mbps goes down, I have to change the static routes to use the 12Mbps link.

What I want is to configure a dynamic routing protocol to converge the traffic to the 12Mbps automatically.

Please, check the attachemt.

Is this possible in this desing?

Or, what should I do to improve the situation..

Thanks

Tauer

[Marwan ALshawi]

try this first

router on LAN 1

ip route 172.16.0.0 255.255.255.0 [the next hop ip on the 34M link]

ip route 172.16.0.0 255.255.255.0 [the next hop ip on the 12M link] 10

look at the 10 at the end

this will increase the route metric

so the first route will be used if the 34M link is down then the router will use the second route which is through the 12M link

the same idea do it on the other side router

good luck

if helpful Rate

[Tauer Drumond]

Hi marwanshawi,

Im using this configuration right now.

But the question is that the two links are a LAN-to-LAN.

Because of this, the router doesn't "understand" that a route is unreacheable.

So, this configuration doesn't work in my scenario.

Another sugestion?

Thanks

[jim.soliven]

Not sure if my previous response made it but here's another try:

Try this:

ip route 172.16.0.0 255.255.255.0 [34M interface] [next-hop of 34M interface]

ip route 172.16.0.0 255.255.255.0 [12M interface] [next-hop of 12M interface] 250

As far as I know, the static route will get installed if it only sees the next-hop route coming from its specified interface.

Hope that helps.

[Tauer Drumond]

Hi,

this kind of configuration still doesn't work.

I'd like to try an OSPF configuration..

Any sugestion?

Thanks

Tauer

[Marwan ALshawi]

ok

i will give this example

LAN1 192.168.0.0/24

router1 34M int: 10.1.1.1/24

router1 12M int: 20.1.1.1/24

LAN2 172.16.0.0/24

router2 34M int: 10.1.1.2/24

router2 12M int: 20.1.1.2/24

Router1 config

router ospf 1

network 10.1.1.0 0.0.0.255 area 0

network 20.1.1.0 0.0.0.255 area 0

network 192.168.0.0 0.0.0.255 area 0

now we gonna configure the ospf cost of both interfaces between the routers manuly to influance the ospf path selection

the lower the prefered

do this on each interface on each router the same

interface [34M int]

ip ospf cost 10

interface [12M int]

ip ospf cost 150

Router2

router ospf 1

network 10.1.1.0 0.0.0.255 area 0

network 20.1.1.0 0.0.0.255 area 0

network 172.16.0.0 0.0.0.255 area 0

now on the interface level on router2 as what we have done on router1

interface [34M int]

ip ospf cost 10

interface [12M int]

ip ospf cost 150

in this case the router will chse the lower interface cost

once the interface with lower cost gos down the other will take over

good luck

please, if helpful Rate

[Tauer Drumond]

Hi marwanshawi,

The problem with a LAN-to-LAN link is that the interface never goes down...

And when a circuit is down, the router doesn't realize this, because the interface still continuous UP and receiving sinalization from DM16E1...

But...with this configuration you sent me, the router can realize else the interface is still UP?

Thanks

Tauer

[Marwan ALshawi]

its ok

even if the interface up but the route will not apear as it will not come from the primary interface then will use the second one

based on the ospf cost

try it and if helpful rate

[Tauer Drumond]

Hi marwanshawi,

I cant do test right now, but ill schedule the change later.

Anyway, I post here the results...

Thanks a lot.

Tauer

[bbaillie]

The links are 34M and 12M that's almost a 3 to one ratio and a kind of waste whenthe 12 meg is idle until afailure occurs. EIGRP is Cisco's favored routing protocol and provides unequal cost load balancing, which means it can take advantage of both pipes and of course keep working when one path fails. The failover is immediate for interface down and 15 seconds for loss of adjacency (more likely you will see a loss of adjacency, carrier equipment most times will keep the interface up, thus making static routes a big problem)

http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a008009437d.shtml

Even better than that is Cisco's newer feature for Cisco express forwarding to distribute traffic further by including the destination address and the port number in the path selection decision. This is a 12.4 T train feature which I suspect will be mainline very soon because it works great.

http://www.cisco.com/en/US/docs/ios/ipswitch/configuration/guide/cef_load_balancng.html

Cheers,

Brian