Why We Need Multi-Areas in OSPF: Internal Router vs ABR vs ASBR

In this lecture, I will discuss multiple areas in OSPF and their importance. OSPF is a highly scalable and efficient routing protocol, but as the size of the network grows, managing it effectively becomes a challenge. This is where the concept of multiple areas comes in.

In OSPF, we have seen up to now that we have Area 0, which is the backbone area. This area is very important because it’s the backbone. You have to think of it as the spine of your body, where all the parts of your body are connected to that spine. Similarly, in OSPF, Area 0is the backbone area, ensuring communication between all other areas in the network.

The Problem with a Single OSPF Area

The problem arises when you have OSPF Area and many routers within it. In such a setup, all routers will send LSAs (Link-State Advertisements), build their link-state databases, and run the SPF (Shortest Path First) algorithm. All of this processing happens within the same area, Area 0. When you have 4, 5, or 10 routers, that’s fine. But as the number of routers grows, the SPF algorithm takes much more processing power to find the best paths within the area. This increased workload consumes significant CPU resources and memory on the routers.

For example, imagine you have 50 routers in a single area. Each router must maintain a detailed view of the network topology, calculate optimal paths, and process changes every time a topology update occurs. This can overwhelm routers, particularly those with limited resources. To prevent this, the network can be divided into multiple OSPF areas.

Normally, you can have up to 80 routers in an OSPF area, but 80 is really the upper limit. I would never recommend having that many routers in one area. Personally, I would start creating another area once I have around 25–30 routers. This approach ensures that the workload on each router remains manageable, and the network can scale effectively.

Benefits of Creating Multiple OSPF Areas

When you create multiple areas, each area becomes its own “domain” for routing processes such as LSAs, the SPF algorithm, and the link-state database. This separation significantly improves the performance and scalability of the network. Let me explain how.

Localized LSA Flooding:
In a multi-area setup, LSAs are confined to their respective areas. For example, changes in Area One will not trigger LSA floods in Area 0. This reduces unnecessary traffic and processing in areas not directly affected by the change.