FHRP with Frame Relay

[gcharan09]

I am trying to enable redundancy protocol on frame relay. Here in my topology I have pair of redundancy routers on both sides of Frame Relay. I know I can configure Router Redundancy for the internal network using any of FHRP, HSRP or GLBP, coming to Frame relay side I have chosen point to point links with sub interface information as follows 

R1 --> R4 as sub interface S0/0.104 

R1 --> R3 as sub interface S0/0.103, similarly for R2, R3 and R4 where R1, R2 act as a redundant pair able to connect to R3, R4 redundant pair.

Now i need help in configuring redundancy for all the data sent from R1 R2 pair to R3 R4 pair and vice versa

Thanks

Charan

[Peter Paluch]

Hi Charan,

I am somewhat confused about what exactly you are trying to accomplish in your topology. FHRP protocols are generally intended to be implemented only in networks with end hosts - they are not supposed to be run between router to provide redundancy, save for some very unusual scenarios.

For providing redundancy between routers, you would normally run any IGP, whether that is RIP, EIGRP, OSPF, or IS-IS. If any of the FHRP protocols was to be used (remember - this is not how they were intended to run), you would need to have your R1, R2, R3, R4 on a single IP subnet which you obviously haven't, as you have written yourself that you are using point-to-point subinterfaces between each pair of routers.

So my general recommendation is: Forget about using FHRP protocols between routers - for this purpose, you are supposed to run a routing protocol instead.

Of course feel welcome to ask further!

Best regards,
Peter

[gcharan09]

Hello Peter Thanks for your comment.

I think I am not clear in explaining my question properly. Even though R1, R2, R3 and R4 are point to point links, they all are in the same subnet.

I wanted to understand the redundancy concept. Lets say we have two routers R1 and R2 which are used for redundancy to route all the traffic of company headquarters. Now lets say we have a remote office of the company with Router R3.

Now i wanted to understand how redundancy works, if R3 wants to send some data to headquarters it can send either to R1 or R2.

How this decision is made and if any of R1 or R2 fails how does R3 knows about it.

Here R1, R2 is connected to R3 via Frame relay.

Hope my question is clear now, sorry for the trouble.

[Peter Paluch]

Hello,

Even though R1, R2, R3 and R4 are point to point links, they all are in the same subnet.

I am not sure if I understand. If you have configured your Frame Relay interconnections using point-to-point subinterfaces, the individual pairs of interconnected routers can not be all in the same IP subnet. That is a violation of the basic principle that if a router has multiple interfaces (or subinterfaces), each of them must be in a different, unique IP subnet.

Perhaps you could post the configurations of R1 - R4 to be more clear on how you are running your network.

Now i wanted to understand how redundancy works, if R3 wants to send some data to headquarters it can send either to R1 or R2.

How this decision is made and if any of R1 or R2 fails how does R3 knows about it.

For this purpose, you would be running any suitable dynamic routing protocol. All routers would run, say, EIGRP. R3 will learn about the headquarter's networks from both R1 and R2 thanks to EIGRP. If the metrics via R1 and R2 are the same, R3 will be using both R1 and R2 in a load-balancing fashion to deliver packets to headquarters. If, say, the metric through R1 is lower than through R2, R3 will be sending packets to headquarters over R1, and only if R1 fails (which R3 will know as it will stop receiving EIGRP Hello packets from R1), it will start using R2.

Does this make sense?

Best regards,
Peter

[mattp0002]

Agree with Peter on this 100%.

Charan-

R1, R2, R3, R(x) etc, should all be routing to/with each other (regardless of the topologies that interconnect them - be it Frame Relay, Ethernet, ATM, Serial lines, IPsec tunnels, or whatever) using a dynamic routing protocol.

Your FHRP (HSRP, VRRP, or GLBP) is only used for router redundancy within one network segment or subnet in order to provide a virtual IP (VIP) address shared between the R1 and R2 routers for the purpose of acting as a default gateway on workstations and other devices which can only be statically configured with one default gateway address.  For example if your headquarters default gateway is 192.168.1.1 and R1 is servicing that IP, a FHRP will allow R2 to take over 192.168.1.1 if R1 goes offline.