cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
20097
Views
41
Helpful
5
Replies

OSPF Point-to-multipoint network type

mandeep62
Level 1
Level 1

Hi Everyone,

 

I'm trying to understand OSPF Network types. One of the type I'm having confusion is Point-to-Multipoint. If I'm correct, this network type emulates a point-to-point connection where no DR/BDR is required and no need to define manual neighbors in order to establish the adjacencies between OSPF routers in NBMA network like frame-relay.

 

How does this Point-to-Multipoint network type work in OSPF in terms of concept?

 

For configuration, I had to enable broadcast flag in frame relay network and hub maintains full adjacencies with spokes.

 

1. Why the addresses 192.168.192.1 and 192.168.192.4 are assigned /32 mask? I have attached the topology.

 

London#show ip route

Gateway of last resort is not set

192.168.192.0/24 is variably subnetted, 3 subnets, 2 masks

C 192.168.192.0/24 is directly connected, Serial0/0

O 192.168.192.1/32 [110/781] via 192.168.192.1, 00:10:04, Serial0/0

O 192.168.192.4/32 [110/845] via 192.168.192.1, 00:10:04, Serial0/0

C 192.168.200.0/24 is directly connected, Loopback0

192.168.232.0/32 is subnetted, 1 subnets

O 192.168.232.1 [110/846] via 192.168.192.1, 00:10:04, Serial0/0

192.168.1.0/32 is subnetted, 1 subnets

O 192.168.1.3 [110/782] via 192.168.192.1, 00:10:04, Serial0/0


 
I would appreciate if someone could explain why we have point-to-multipoint and point-to-multipoint non-broadcast in ospf and where do we use these types.
 
Thanks.
1 Accepted Solution

Accepted Solutions

Peter Paluch
Cisco Employee
Cisco Employee

Hello Mandeep,

The Point-to-Multipoint (PtMP) network type in OSPF is ideal for network segments that do not provide full mutual connectivity between the attached routers. The main characteristics of this network type are:

  • Multicasts are still used to allow the neighbors to detect each other dynamically
  • There is no DR or BDR elected
  • Each router will advertise its own IP address on the PtMP network with a netmask of /32
  • The link-state database will represent the connectivity between routers as it really exists, based on which routers can hear each other

Especially the last property is important: With Broadcast or Non Broadcast Multi Access network types, OSPF implicitly assumes that on the network segment, every router can hear each other even if that is not really true. In that case, the network needs quite some tweaking to help OSPF avoid making bad decisions based on this assumption. Because of this property, the PtMP network type is ideal for hub-and-spoke topologies or partially meshed topologies that might be seen with Frame Relay, ATM, but also with Carrier Ethernet's E-Tree service.

The next-to-last property - each router advertising its own IP address with a /32 netmask - is used to fix the reachability between the neighbors on the network that do not have a direct visibility to each other (for example, between spokes). OSPF will treat that IP address as any other, and find out the shortest path to it along the sequence neighbors that can talk to each other directly (from one spoke to hub and from there to the destination spoke).

Point-to-Multipont Non Broadcast (PtMP-NB) network type is a slight twist on the PtMP network type from which it borrows all the properties, with one particular difference: The neighbors have to be configured manually, and use unicast to talk to each other. The PtMP-NB network type is the most generic OSPF network type, since it gives you an absolute control which routers on a common segment will form adjacencies; it even allows you to configure the cost of the adjacency between each neighbors, so even though all neighbors are connected to the same network, some of them may be configured to appear "closer" to each other than the others.

Please feel welcome to ask further!

Best regards,
Peter

 

View solution in original post

5 Replies 5

balaji.bandi
Hall of Fame
Hall of Fame

Peter Paluch
Cisco Employee
Cisco Employee

Hello Mandeep,

The Point-to-Multipoint (PtMP) network type in OSPF is ideal for network segments that do not provide full mutual connectivity between the attached routers. The main characteristics of this network type are:

  • Multicasts are still used to allow the neighbors to detect each other dynamically
  • There is no DR or BDR elected
  • Each router will advertise its own IP address on the PtMP network with a netmask of /32
  • The link-state database will represent the connectivity between routers as it really exists, based on which routers can hear each other

Especially the last property is important: With Broadcast or Non Broadcast Multi Access network types, OSPF implicitly assumes that on the network segment, every router can hear each other even if that is not really true. In that case, the network needs quite some tweaking to help OSPF avoid making bad decisions based on this assumption. Because of this property, the PtMP network type is ideal for hub-and-spoke topologies or partially meshed topologies that might be seen with Frame Relay, ATM, but also with Carrier Ethernet's E-Tree service.

The next-to-last property - each router advertising its own IP address with a /32 netmask - is used to fix the reachability between the neighbors on the network that do not have a direct visibility to each other (for example, between spokes). OSPF will treat that IP address as any other, and find out the shortest path to it along the sequence neighbors that can talk to each other directly (from one spoke to hub and from there to the destination spoke).

Point-to-Multipont Non Broadcast (PtMP-NB) network type is a slight twist on the PtMP network type from which it borrows all the properties, with one particular difference: The neighbors have to be configured manually, and use unicast to talk to each other. The PtMP-NB network type is the most generic OSPF network type, since it gives you an absolute control which routers on a common segment will form adjacencies; it even allows you to configure the cost of the adjacency between each neighbors, so even though all neighbors are connected to the same network, some of them may be configured to appear "closer" to each other than the others.

Please feel welcome to ask further!

Best regards,
Peter

 

Hi Peter,

Thank you for replying. To sum it up, OSPF on a shared network like frame relay (Hub and Spoke) where the default underlying nature of network is NBMA. The workaround for the neighbor relationship is to manually define the neighbors for the adjacencies to come up, and make the hub as DR if we want to use only NBMA which is default for frame relay. The packets are sent as unicasts as NBMA doesn't allow replication of broadcast/multicasts.

Now, if we want our NBMA network to be seen as a collection of point to point links in terms of OSPF, we can change the network type to Point to Multipoint in OSPF which as you said is ideal for hub and spoke or partially meshed topologies. In a hub and spoke topology, spokes establish FULL adjacency with the hub only and exchange their LSAs with the hub which in turn forwards to the other spoke.

However, I'm trying to understand the database of the topology which I'm having a confusion about number of the links of Router 2 which is only 2.

 

ospf_Net.png

 

Here is the output of

sh ip ospf

database router

 

R1#show ip ospf database router

            OSPF Router with ID (1.1.1.1) (Process ID 1)

        Router Link States (Area 0)

  LS age: 1454
  Options: (No TOS-capability, DC)
  LS Type: Router Links
  Link State ID: 1.1.1.1
  Advertising Router: 1.1.1.1
  LS Seq Number: 80000010
  Checksum: 0x61B2
  Length: 72
  Number of Links: 4

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 1.1.1.1
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 1

    Link connected to: another Router (point-to-point)
     (Link ID) Neighboring Router ID: 3.3.3.3
     (Link Data) Router Interface address: 192.168.1.1
      Number of TOS metrics: 0
       TOS 0 Metrics: 64

    Link connected to: another Router (point-to-point)
     (Link ID) Neighboring Router ID: 2.2.2.2
     (Link Data) Router Interface address: 192.168.1.1
      Number of TOS metrics: 0
       TOS 0 Metrics: 64

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 192.168.1.1
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 0


  LS age: 346
  Options: (No TOS-capability, DC)
  LS Type: Router Links
  Link State ID: 2.2.2.2
  Advertising Router: 2.2.2.2
  LS Seq Number: 80000005
  Checksum: 0x569D
  Length: 48
  Number of Links: 2

    Link connected to: another Router (point-to-point)
     (Link ID) Neighboring Router ID: 1.1.1.1
     (Link Data) Router Interface address: 192.168.1.2
      Number of TOS metrics: 0
       TOS 0 Metrics: 64

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 192.168.1.2
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 0


  LS age: 310
  Options: (No TOS-capability, DC)
  LS Type: Router Links
  Link State ID: 3.3.3.3
  Advertising Router: 3.3.3.3
  LS Seq Number: 80000006
  Checksum: 0x9536
  Length: 60
  Number of Links: 3

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 3.3.3.3
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 1

    Link connected to: another Router (point-to-point)
     (Link ID) Neighboring Router ID: 1.1.1.1
     (Link Data) Router Interface address: 192.168.1.3
      Number of TOS metrics: 0
       TOS 0 Metrics: 64

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 192.168.1.3
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 0

 

Why is the number of links 2 by router2 in other words a stub network of it's own loopback 2.2.2.2 is missing?

 

Thanks for the help again.

Hello Mandeep,

You are welcome!

Your sum-up is very nice and touches all the important points nicely. One thing to keep in mind: With NBMA network type, the DR router makes a leap of faith in two aspects:

  1. It assumes that whatever neighbors it can hear, they are the complete set of neighbors; there is no possibility for a valid neighbor to go undetected
  2. It assumes that all neighbors of the DR can also hear and talk to themselves directly. This is then translated into the contents of the link state database.

Hub-and-spoke topologies break both these points. Spokes cannot hear each other, so if a spoke became a DR, it would have a very distorted view of the network - it would know only about itself and the hub. Also, a spoke cannot send packets to another spoke directly, even if the link state database suggests that.

The PtMP network type does not do any of these assumptions. It meticulously assumes connectivity only between those neighbors that can hear each other's Hellos and can become fully adjacent. As a result, the link state database contains a faithful representation of the real connectivity on the segment.

Why is the number of links 2 by router2 in other words a stub network of it's own loopback 2.2.2.2 is missing?

Most likely, the loopback is either not up/up, or it is not added to OSPF. Can you double-check this? The

show ip ospf interface brief

is your friend - it gives you a quick overview of all OSPF-enabled interfaces.

Best regards,
Peter

Thank you, Peter. You were right, Loopback was not added. Thank you again. Cheers.
Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Review Cisco Networking products for a $25 gift card