OSPF Design: Stub Area Types

I initially started this blog to share my CCDE journey, what I used to prepare, what’s next, etc., but I also want to write a series of random technical reviews that align with the CCDE Written Exam Topics. This will serve as a good exercise for me as I prepare for the v2 Written Exam, and hopefully others will find it somewhat useful. The topics will be random, usually when I want to expand on a question that I’ve tried to answer in a forum, on the job, or something that I’ve run into in my studies.

Let’s start with a review of OSPF stub area types –

Stub Area
Redistributed routes a.k.a. OSPF external routes are not advertised into a stub area. OSPF inter-area routes are advertised into the area, and the ABR will inject a default route into a stub area.

Totally Stubby Area
Totally Stubby takes it a step further… OSPF external routes and inter-area routes are not advertised into a Totally Stubby Area. The ABR injects a default route into a Totally Stubby Area.

Not So Stubby Area (NSSA)
This is where it gets somewhat tricky… OSPF external routes are not advertised into a Not So Stubby Area (NSSA) just like in a stub area. OSPF inter-area routes are advertised into an NSSA just like in a stub area. By default, the ABR will not inject a default route into the NSSA. The big difference though is that NSSA allows you to redistribute external routing information into the area. The ASBR generates a Type 7 LSA that contains the external routing information. The ABR receives each Type 7 LSA, translates it to a Type 5, and floods it into area 0.

Totally Not So Stubby Area (Totally NSSA)
Just like in a Totally Stubby Area, OSPF external routes and inter-area routes are not advertised into a Totally Not So Stubby Area (Totally NSSA). In this case the ABR will inject a default route into the Totally NSSA. Just like in NSSA, the ASBR generates a Type 7 LSA that contains the external routing information. The ABR receives each Type 7 LSA, translates it to a Type 5, and floods it into area 0.

Why is this important? The general design rule to follow with OSPF is to “separate complexity from complexity.” For example, you would want to separate a large hub and spoke topology from the rest of the network, so you would place the hub and spoke topology in a separate area or flooding domain. A failure of one of the spoke routers doesn’t impact the rest of the network outside of the area. Hiding or limiting topology information makes the network more stable, helps provide faster convergence, and allows for scalable OSPF routing design. Stub areas optimize this concept of reducing flooding and they effectively limit the size of the flooding domain.

Advertisements