Hi,

the two central site routers have seperate bri (or pri) interfaces for this connection with separate numbers.

Regards,

Thorsten

Hi,

With two separate numbers it's a little bit more complicated. You will have no other choice than configuring two dialer strings in the branch routers. It will attempt the first, then if it fails, will attempt the second one. You can configure the branch routers in two sets with the order of the strings inverted,to have some load balance.

There is a minor chance that the PPP multilink will terminate calls on the two central routers, however I would personally live with that rather than adding the complexity of configuring multichassis multilink, that is not well suited for enterprise networks.

On the central router the interfaces will be in the same subnet but with different IP addresses. Then you will configure floating static routes, and define all routing traffic as non interesting to the effect of keeping up the link. This will cause that once the primary circuit comes back, the ISDN link will be torn down.

Generally for simplicity and less overhead, RIp routing with filters is configured on hub and spoke networks. That does not prevent you from running OSPF in the core,with the appropriate redistribution.

Hope this helps, please rate post if it does!

Hi Paolo,

thanks for your answer, which nearly completely solved my problem.

Is it correct that both central routers and the branch router are connected to the same ip subnet?

E.g. the subnet is 10.1.1.0/24.

central router 1: 10.1.1.1

central router 2: 10.1.1.2

branch router: 10.1.1.10

configuration of the branch router loooking like:

...

interface BRI0/0

no ip address

encapsulation ppp

no ip mroute-cache

dialer pool-member 1

isdn switch-type basic-net3

!

interface Dialer1

bandwidth 128

ip address 10.1.1.10 255.255.255.0

encapsulation ppp

no ip mroute-cache

dialer pool 1

dialer string

dialer string

dialer hold-queue 100

dialer load-threshold 1 either

dialer-group 1

no cdp enable

ppp authentication chap

ppp multilink

...

Regards,

Thorsten

Yes. What are you missing is the dialer-list, that define which traffic is condired interesting in order to keep up the link. For this you specify an access-list that only permits, for example, unicast source from the branch LAn and nothing else. See:

http://cisco.com/en/US/docs/ios/12_0/dial/command/reference/drhowddr.html

Hope this helps, please rate post if it does!

Yes,

I already configured it but didn't post it.

I will test this design and give a short feedback about the practical result.

Regards,

Thorsten

The problem with 'backup' command is thats it doesn't make the line go active when for example there is a 'black hole' in the network, or the FR PVC status is not reliable, or many other situations that dictates the use of floating static routes for a professional design.