Frame relay P2P and Multipoint

gurkamal01 · ‎06-14-2008

Hi

I was reading about frame relay in a ccna book and i am using 2 ccna books to learn about it

In one book it says that P2P frame relay is best suited for partially meshed networks and require a seperate subnet number whereas multipoint is suited for fully meshed

The other book shows the lab example of a hub and spoke topology where R1 is the HUB and R2 and R3 are the spokes

R2 interface is configured with a multipoint subinterface with two frame map commands to R1 and R3 each having the same dlci

since dlci are locally significant how can we use the same dlci no. on both the frame map commands on R2

R1 ip add on s0 is 172.12.123.1 on r2 ip add on s0 is 172.12.123.2 and R3 add is 172.12.123.3

my second concern is R3 uses a point2point subinterface but all the ip address configured on the serial interfaces of R1,R2 and R3 are in the same subnet how can we use point2point subinterface on R3 as it states p2p require a different subnet

my 2 questions are

1. why same dlci no. are used on r2 which has a multipoint subinterface on the frame map command

2. Is it necessary to use different subnet number when configuring P2P subinterface on R3

3.why do we use multipoint and P2p in frame relay

Please help

Thanks

sdoremus33 · ‎06-14-2008

The difference between Frame For Physical and PTOMP environements (Hub Spoke) you would use the frame-relay map command (See Below)

lets say your ip address id x.x.x.x and dlci is y

Frame relay map ip x.x.x.x y broadcast, where x.x.x.x = ip address and y is the remote dlci number associated.

Note:remember to use the broadcast keyword

Basically PTOMP treats each spoke link in the same broadcast network (subnet)

In PTOP network you would use the frame-relay dlci command

PTOP networks basically act as seperate point-to-point subnet links.HTH

One reason we use PTOMP is because it saves on address consumption by using the same broadcast subnet

gurkamal01 · ‎06-14-2008

Hi

I know that when configuring PTOMP we use the frame map ip command and when configuring PTOP we use the frame-relay interface dlci command

but my questions are

1. Is when configuring PTOP links is it necessary to use different subnet #s ?

2. According 2 the lab scenario which i mentioned earlier in the post, why do we have to use the same dlci # for the 2 VC's from R2 (one VC to R1 hub and the other VC to R3) as dlci have local significance?

I think it should be two different dlci #s from R2.

Appreciate any help.

hennigan · ‎06-14-2008

1. Because R2 is a spoke, it is connected directly only to R1 via one PVC. You need one map statement pointing to R1 which is directly connected. The second map statement pointing to R3 out the same DLCI provides reachability to R3 via R1. Traffic from R2 to R3 travels first to R1, then is routed via R1 out to the PVC between R1 and R3.

2. P2P frame-relay links operate much like conventional P2P serial lines on different physical wires. As such, each link is in a separate subnet.

3. Multipoint conserves IP addresses as many PVCs can share the same subnet. P2P allows each PVC to be treated more like a separate physical wire. Multiple P2P links simplify the configuration in a hub-and-spoke scenario as each spoke doesn't need to be mapped to the others specifically.

gurkamal01 · ‎06-14-2008

hi

thank you so much

As you said "The second map statement pointing to R3 out the same DLCI provides reachability to R3 via R1"

so we can use the same dlci number on the second frame map ip command pointing to R3 going via R1

Thanks

hennigan · ‎06-14-2008

Exactly. R2 is a spoke router. It only has one PVC attached to it, from R1. The second map statement pushes traffic destined for R2's IP address out that PVC. R1 receives the traffic and routes it to R3.

If you had another PVC connecting R2 directly to R3, then you would map R3's address to that direct DLCI. This would be a full mesh (triangle) between the three routers.

In production networks, you will typically see only P2P links within a frame-cloud on a hub-and-spoke scenario as it makes the configuration easier. You'll find a mix of both multipoint and P2P a lot more often on Cisco exams than in the real world.

With multipoint on the spokes, it becomes necessary to add the map to every spoke whenever a new spoke is added.

gurkamal01 · ‎06-14-2008

hi

As you said

"The second map statement pushes traffic destined for R2's IP address out that PVC. R1 receives the traffic and routes it to R2.

should it be "R3's ip address out that pvc" in the above statement and not R2's if i am not wrong its a typo coz R2 has the the two frame map ip commands on its subinterface whereas R3 has a ptop link to R1

Appreciate your responses

gurkamal01 · ‎06-14-2008

Hi

I am just on the verge of marking this conversation as resolved but can you please reply to my confusion and let me know if it is a typo or not

Thanks

hennigan · ‎06-14-2008

It's a typo. The second map on R2 maps R3's IP to the PVC connected directly to R1. R1 routes the traffic to R3.

hennigan · ‎06-14-2008

Yes, it is the only DLCI appearing on that interface, so you need to map both R1 and R3 to it.

gurkamal01 · ‎06-14-2008

hi

thank you so much

As you said "The second map statement pointing to R3 out the same DLCI provides reachability to R3 via R1"

so we can use the same dlci number on the second frame map ip command pointing to R3 going via R1

Thanks

gurkamal01 · ‎06-14-2008

Hi

Am i right with that???

Thanks

Thotsaphon Lueangwattanaphong · ‎06-14-2008

gursaran,

You are right. You need Layer3-to-Layer2 mapping to send it out on the same DLCI connecting to the R1. Because you only have 1 PVC.

HTH

Thot

gurkamal01 · ‎06-15-2008

HI

As above mentioned,is this the way it works on hub and spoke enviroment or every other topology

1.How is frame relay used in production enviroment hub and spoke or something else?

can anyone please inform me about that

2.can we add another pvc from R2 to other routers in a home lab enviroment or a real world enviroment in hub an spoke topology or any other toppology

If yes that in which scenario and how??

please advice

Thotsaphon Lueangwattanaphong · ‎06-15-2008

gursaran,

1.Hub and Spoke is fine to be used in real world. It's very likely to be Point-to-Point on Hub and Spokes. It's easy to configure. You need to separate subnet on each link.

2. You can add, That's virtual-circuits to connect to the other routers. If you add that then shouldn't be Hub and Spoke design any more.

HTH

Thot