This is the fifth section of the OSPF series of articles. As before, we will be working with the following diagram:

OSPF-DIAGRAM-1

We will pick up from where we left off and that is OSPF area types. We covered both the stub and totally stubby area so far which leaves us with the following two area types:

Not-So-Stubby Area

Totally Not-So-Stubby Area

In order for us to understand these two area types let’s suppose that area 45 which is in a different geographical location absorbed another company. This would look like the following in our diagram:

OSPF-DIAGRAM-2

We have RIP version 2 being used between R5 and R7 where R7 is the router that connects the company that we just absorbed to our network. This is a temporary situation until we can fully integrate the network from the other company into ours. We will configure this scenario and then redistribute the external RIP routes into our OSPF network so that our other locations for our enterprise can communicate with the company that we just absorbed. You want to start off by assigning IP addresses to the link between R5 and R7 as we have done before. You can then configure the RIP routing protocol on both R5 and R7 which we covered in a previous section[INSERT LINK]. Here is what the configuration on R7 looks like:

R7(config)#router rip
R7(config-router)#version 2
R7(config-router)#passive-interface Loopback0
R7(config-router)#passive-interface Loopback1
R7(config-router)#network 10.0.0.0
R7(config-router)#network 172.57.0.0
R7(config-router)#no auto-summary

Note: I am simulating the 172.57.x.x networks by using loopback interfaces.

Once we have configured R5 then this is what we see on the routing table.
OSPF-ROUTE-1

Noticed that there are two new routes that have been learned on this router and they are both RIP routes. Next we will advertise these routes into the OSPF routing process using the following command on R5.

R5(config)#router ospf 25
R5(config-router)#redistribute rip subnets

Redistribution is an advanced routing topic that I will not cover or go into detail here. The results of the command is that the other routers in our OSPF environment will now know about the two RIP routes. Here is the routing table on R4:
OSPF-ROUTE-2

Notice the two 172.57.x.x networks being learned from R5. These two routes will propagate throughout our entire OSPF infrastructure. Now that we have gotten this far we can start to introduce the Not-So-Stubby Area. From the previous article we know that the purpose of a stub area is to filter external routes from entering an OSPF area. By doing so we will be decreasing the size of the routing table which leads to faster routing decisions being made by the router. Area 45 can’t be made into a stub area because it has an ASBR(R5) that is injecting external routes learned via RIP into OSPF. The solution to this dilemma is a Not-So-Stubby Area which will allow us to filter external routes from other areas and still allow R5 to send external routing information for redistribution into OSPF.

Like before, when we configure an area to be Not-So-Stubby we must change the configuration on the ABR and all the routers within that area. In R4 we would have to apply the following command under the OSPF routing process.

R4(config)#router ospf 25
R4(config-router)#area 45 nssa default-information-originate

In R5 we can use the same the command with the default-information-originate portion which I have explained below.

Let’s explain what is happening behind the scenes so that we have an understanding of what is occurring here. An NSSA(Not-So-Stubby Area) will block external routes from entering the area by filtering external LSAs(Type 4 ASBR Summary). At the same time we have an ASBR(R5) injecting external routes which would also get blocked otherwise. The solution to this is that OSPF will use a new type 7 LSA just for the external routes that R5 is advertising. The type 7 LSA will be advertised by R5 to R4 and R4 will then advertised a type 5 LSA to area 0. Let’s look at the routing table on R5 to see what is happening.

Note: The default-information-originate portion in the command above allows R4 to inject a Type 7 LSA default route so that R5 can point to R4 to have that as its gateway of last resort.

OSPF-ROUTE-3

Note the N2 type route at the bottom which stands for OSPF NSSA external type 2 and is a type 7 LSA. This is what R4 is passing to R5 so that R5 is capable of reaching external networks such as the EIGRP learned routes via ASBR R1. If we look at the routing table for R4 then we will see the following:

OSPF-ROUTE-4

Notice that R4 is also learning the N2 type routes via R5 which are type 7 LSAs. R1 as we explained should not have any N2 routes since R4 is taking the type 7 LSAs and advertising them into area 0 as type 5 LSAs.

OSPF-ROUTE-5

Note that R1 is learning all of these routes as E2 and not N2. Let’s take a quick look and see what wireshark picks up when R5 advertises a type 7 LSA to R4.

OSPF-WS-1

We can see that these networks are all getting advertised using type 7 LSAs to R4. On the other hand looking at the packet capture on the interface that R4 connects with to area 0 we can see that these are getting advertised as Type 5 LSAs into area 0.

OSPF-WS-2

So far we have covered three area types and we have one more remaining. The last area type is the totally NSSA. From the previous section about the totally stubby area we know that when we made an area into a totally stub that the inter area routes were also filtered. A Totally NSSA does the same thing in that it will filter inter area routes as well as what a regular NSSA does. Let’s go ahead and make this change on our routers in area 45 to see what the results are.

R4(config-router)# area 45 nssa no-summary

The command is the same except that we are using the no-summary post fix. If we look at the routing table on R5 then we notice that the following has occurred.

OSPF-ROUTE-6

There are no external routes and only one inter area route. All of those routes have been replace by a single OSPF candidate default inter area route pointing to our ABR(R4). Everything else is the same and what I explained above for a regular NSSA still applies. Before we finish this article we will do some verification commands.

Let’s start of by doing a trace route from R6 to R7 to see if our implementation is working properly.

TRACEROUTER

We successfully reached our destination and we traverse R1, R4, and R5 to reach the networks behind R7.

The majority of OSPF commands have been covered in previous sections but here are a few that I did not mention yet. The first one is the “show ip ospf database router x.x.x.x” command which will show us more details about type 1 router LSAs which we discussed previously. E.g. below is the type 1 LSA that R4 knows about from R1.

OSPF-DATABASE-1

Similarly if we want to see more details on type 2 or type 3 LSAs then the following commands will give us this information:

R4#show ip ospf database network

R4#show ip ospf database summary

OSPF-DATABASE-2

OSPF-DATABASE-3

The “show ip ospf database TYPE” command also gives you the option to view other link state types. I listed the common ones above since these are the ones that you will encounter the most. This is the final article in the OSPF series and I hope that the information provided was useful. Thank you for taking your time to read this article. Happy holidays and see you here next time.

Leave a Reply