Introduction to Single-Area OSPF and Multi-Area OSPF
Because of its hierarchical design, OSPF is more efficient and scalable. The OSPF segments the network into different areas. An area is a group of routers sharing the same link-state information in their Link-State Databases (LSDBs). We can implement the OSPF in two ways.
- Single Area OSPF
- Multi-Area OSPF
Single Area OSPF (Area 0)
Area 0 is also known as the backbone area for OSPF, which links all other smaller areas within the hierarchy. The single area OSPF is helpful in smaller networks where only a few routers are working, the web of router links is not complex, and paths to individual destinations are easy. If an area becomes too big, the network will be affected.
- Large routing table
- Extensive link-state database (LSDB)
- Frequent SPF algorithm calculations
OSPF supports hierarchical routing using areas, which makes OSPF more efficient and scalable. An area is a group of routers sharing the same link-state information in their databases. The figure below illustrates the single-area OSPF network.
Multi-Area OSPF
When a large OSPF area is separated into several smaller areas, this is called multi-area OSPF. It is helpful in larger network deployments to decrease processing and memory overhead. The network contains OSPF areas in a hierarchal design. All areas are connected to area 0, known as the backbone area—the interconnecting routers between the backbone area and other areas are known as area border routers (ABR).
The figure shows that multi-area OSPF segments one large network into several networks. One autonomous system ( AS) has been divided into multiple areas to support hierarchical routing. The routing occurs between the multiple areas. The more operations that require system resources, such as recalculating the database, are done only within a relevant area.
For example, when changes occur in the topology, like addition, deletion, or modification of the link, the router must rerun the SPD algorithm to create a new SPF tree and update the routing table in the same area where changes occur. However, changes have been shared with routers in other areas in a distance-vector format to update their routing tables, and these routers and areas do not need to rerun the SPF algorithm. As illustrated in Figure 1, the hierarchical-topology possibilities of multi-area OSPF have these advantages:
- Smaller routing tables—Fewer routing table entries as network addresses can be summarized between areas. For example, R1 summarizes the routes from Area 1 to Area 0, and R2 summarizes the routes from Area 51 to Area 0. R1 and R2 also propagate a default static route to Area 1 and 51.
- Reduced link-state update overhead– Minimizes processing and memory requirements because fewer routers exchange LSAs.
- Reduced frequency of SPF calculations—This Minimizes the impact of a topology change within an area. For instance, it minimizes the impact of routing updates because LSA flooding stops at the area boundary.
- Small Link-State Data Base. Multi-areas OSPF efficiently partitions a potentially large database into smaller, more manageable ones.
OSPF Two-Layer Area Hierarchy
Multiarea OSPF is implemented in a two-layer area hierarchy:
- Backbone (Transit) area—An OSPF area transfers IP packets fast and efficiently. Backbone areas interconnect multiple OSPF areas. End users are usually not found within a backbone area.
- Regular (Non-backbone) area –This area connects users and other resources. The areas are usually set up along functional or geographical groupings. By default, a regular area does not allow traffic from another area. Traffic from other areas must forward through a transit area. A regular can contain several subtypes, including a standard area, stub area, totally stubby area, and not-so-stubby area (NSSA). The Cisco recommends the following guidelines for areas:
- The maximum number of routers per area is 50.
- One router can join a maximum of three areas.
- The maximum number of neighbors for a single router is 60.