Learn Cisco

Cisco IOS OSPF modes:

1. Point-to-point (Cisco)
2. Broadcast (Cisco)
3. Non-broadcast (RFC 2328)
4. Point-to-multipoint (RFC 2328)
5. Point-to-multipoint non-broadcast (Cisco)

Topology

Cisco IOS OSPF modes:

1. Point-to-point (Cisco)
2. Broadcast (Cisco)
3. Non-broadcast (RFC 2328)
4. Point-to-multipoint (RFC 2328)
5. Point-to-multipoint non-broadcast (Cisco)

Topology

Cisco IOS OSPF modes of Operation:

1. Point-to-point (Cisco)
2. Broadcast (Cisco)
3. Non-broadcast (RFC 2328)
4. Point-to-multipoint (RFC 2328)
5. Point-to-multipoint non-broadcast (Cisco)

Topology

Icons designed by: Andrzej Szoblik - http://www.newo.pl

Workflow

1. Get familiar with the topology.
2. Read the 'Task List'.
3. Read the 'Questions' and provide the answers BEFORE configuring the routers.
4. Configure the lab according to the 'Task List' .
5. Compare the answers (step 3) with the lab results.

Task List

• Configure IP addresses as per Pic. 1.
• Configure Frame-Relay using DLCIs listed in the topology (Pic. 1). Make sure that S0/0 interfaces do NOT learn any other DLCIs.
• Enable OSPF in all routers. All interfaces should be in area 0. Use default OSPF mode.
• Check if OSPF adjacency has been built and if the loopback interfaces are reachable.

Questions

1. What is the default OSPF mode of operation on NBMA networks (here: Frame-Relay)?
2. Is DR/BDR elected by default in NBMA networks?
3. If DR/BDR are elected, which router in the topology can be DR and which can be BDR?
4. Do OSPF packets use multicast or unicast transmissions by default?
5. What kind of well-known issue with regards to IP-to-DLCI mapping can a router run into? How can I check it? How can I fix it?
6. If Frame-Relay hub-and-spoke topology is used, what are the differences between OSPF non-broadcast mode and OSPF point-to-multipoint mode?
7. R1's interface S0/0 (Frame-Relay) uses the default OSPF mode, and R2's interface S0/0 (Frame-Relay) is configured as multipoint interface (S0/0.1 multipoint). What OSPF mode is going to be used on S0/0.1?
8. With hub-and-spoke topology (the spokes do not have circuit provisioned) what is the behavior of the next-hop address between the spokes?

Lab Solution

Note!
Since Frame-Relay is not a full meshed, the spoke routers must NOT become DR or BDR! 

Note!
When static mapping IP-to-DLCI is uses the broadcast keyword allows broadcast and multicast to be sent out as unicast (pseudo broadcast/multicast).

R1 Configuration:

R2 Configuration:

R3 Configuration:

Verification

Note!
R1 (the hub router) uses the 'neighbor' statement to send hello packet using unicast. Without this command, the adjacency will not get established.

Note!
R2 and R3 are DROTERs as they must NOT become DR/BDR since they have no full reachability to other routers in the topology.

Note!
Frame-Relay static mapping on R2 also shows how to reach R3 via R1 (no broadcast keyword is necessary since no broadcast/multicast are gonna be sent between R2 and R3 directly).

Quick Note:
I must keep my Video Notes as short as possible since Youtube upload takes a moment. Each post is going to have only two videos: the answer to the question, and the lab solution.

Cisco IOS OSPF modes of Operation:

1. Point-to-point (Cisco)
2. Broadcast (Cisco)
3. Non-broadcast (RFC 2328)
4. Point-to-multipoint (RFC 2328)
5. Point-to-multipoint non-broadcast  (Cisco)

In this post I'm going to refresh few concepts with regards to OSPF operation in broadcast networks.

Again, the tasks helping me understand the OSPF operation are going to be structured in a similar way to my previous post (OSPF in point-to-point networks)

Topology

Icons designed by: Andrzej Szoblik - http://www.newo.pl

Workflow

1. Get familiar with the topology. Pay attention to what Router IDs are on each router.
2. Read the 'Task List'.
3. Read the 'Questions' and provide the answers BEFORE configuring the routers.
4. Configure the lab as per 'Task List' .
5. Compare the answers (step 3) with the lab results.

Task List

• Configure IP adresses as per topology diagram (pic. 1). Check the layer 3 connectivity.
• Enable OSPF routing protocol. All interfaces should be in area 0.
• R1 should be elected the DR and R2 should be elected the BDR on the broadcast segment.
• Make sure that loopback interfaces are advertised with their configured network mask (/24).
• On each router use different method of enabling OSPF on the interfaces.

 Questions

1. What is the default OSPF mode of operation on broadcast links?
2. What are the default timers (hello/dead) used on broadcast links?
3. What is the role of DR/BDR on broadcast/NBMA segment (more detailed answer can be found in my other blog here)?
4. If all routers participate in the election of DR/BDR what parameter(s) determines these roles?
5. If the DR/BDR have been elected, what is going to happen if the router with the highest router ID joins the broadcast/NBMA network?
6. What is going to happen if DR is no longer available and BDR has lower priority than the newly introduced router?
7. What is the implication of using the priority with the value of 0?
8. If R1 is the DR and R2 is the BDR, what is going to be OSPF state between R3 and R4 in the topology used (pic.1)?
9. What is the address used to advertise updates on broadcast segment?

Lab Solution

Note!
Since the lab stipulates that R1 must be the DR and R2 the BDR, the order of operation does matter (unless I want to clear the ip ospf process later which is a waste of time). That's why I start my configuration with R1 and R2 first by changing the priority on F1/0 interfaces. 

Note!
Because the loopback interfaces area advertised as stub network (/32) by default, I change the mode of operation to point-to-point so OSPF advertises them with their native (configured /24) network mask. 

Note!
I am going to use different ways of enabling OSPF on the interfaces. In practice (especially in the CCIE lab), the most specific wildcard mask is recommended unless specified otherwise.

R1 Configuration:

R2 Configuration:

R3 Configuration:

R4 Configuration:

Verification

Note!
On R4 OSPF neighbor adjacencies. Pay attention, R4 has 2-way (correct) state with R3. Full adjacency established between R1 (DR) and R2 (BDR).

Note!
In my configuration R4 redistributes connected network (Lo0) into OSPF. This shows as E2 (external type 2) by default.

Note!
R4 learns all subnets (loopbacks advertised by R1, R2, and R3) with /24 network mask. This is the result of using point-to-point mode of operation configured on the loopback interfaces.


Note!
R1 (also R2 and R4) learns 172.16.104.0/24 as external (redistributed) subnet.

---

This post contains the answers and configuration for the LAB 1.

Answers
Q1:
Answer:
The default OSPF mode of operation is 'broadcast'.

Q2:
Are hello packets sent using unicast or multicast? 
Answer:
Hello packets will use multicast 224.0.0.5.

Q3:
Answer:
Yes, since broadcast and NBMA interfaces do participate in DR/BDR election unless have been configured with the priority=0.
The reason for electing DR/BDR on broadcast and NBMA networks is to limit the propagation of LSAs. DRother routers establish full adjacency with DR and BDR only. DRothers establish 2-way adjacency with other routers in DRother role.

Q4:
Answer:
The default encapsulation on S0/1 is HDLC which is point-to-point protocol.
The default OSPF mode is point-to-point. This is the default OSPF mode on the following:

• Interfaces running point-to-point protocols (HDLC, PPP)
• Subinterfaces configured as point-to-point (int s0/1.point-to-point) 

Q5:
Are hello packets sent using unicast or multicast?
Answer:

Since point-to-point connections support multicast (like broadcast networks), multicast 224.0.0.5 is going to be used when sending hello packets. 

Q6:
Does it participate in DR/BDR election? Why?
Answer:
OSPF point-to-point mode does not participate in the election of DR/BDR since there are only two routers echanging LSAs. There is no excess of LSA exchanges typical for multi-access networks.

Q7:
What are the default hello and dead interval timers used on F1/0 and S0/1?
Answer:
On both interfaces the timers are:

• hello=10 seconds
• dead=40 seconds. 

Q8:
Is there any other OSPF mode of operation that could work with the default OSPF mode on F1/0? What would you have to do to make it work?
Answer:
The default OSPF mode on broadcast networks is broadcast.
You can configure OSPF-enabled interface with NBMA mode which also elect DR/BDR. In order to make them work together you have to adjust the timers as they are not same between these two modes. NBMA mode uses:

• hello=30 seconds
• dead=120 seconds

Q9:
Answer:
The default OSPF mode on S0/1 (HDLC encapsulation) is point-to-point.
The "somewhat" compatible modes of operation with the 'point-to-point' are:

• point-to-multipoint
• point-to-multipoint nonbrodcast

Also, their hello and dead interval timers must be tweaked. They use 30/120 whereas point-to-point use 10/40.

Q10:
Answers:
Quick solutions could be:

1. The interface command: 'ip ospf 1 area 0' (here 1 is the process ID).
2. Redistribution of connected networks into OSPF.

As for the solution to the tasks, here goes...

R1 Configuration:

R3 Configuration: