Overview

Proper use of OSPFv3 Areas can greatly reduce the complexity of OSPFv3. Benefits include reducing the demand on network resources and increasing scalability.

The hierarchical use of OSPFv3 Areas make the protocol very scalable. Areas are a logical grouping of links.

All routers in an Area, by definition, have identical Link State Databases (LSDBs). All routers in the same Area:

  • Receive the same link state information
  • Have the same result from OSPF’s Dijkstra algorithm for each destination’s shortest path

With the use of Areas, instead of a router needing to maintain a routing table of all routers it can simply have summary routes to other Areas.

Point-to-Point Topology

Let's begin by taking a look at a topology of a simple point-to-point link:

Point to Point Topology
Point to Point Topology

This example is straight forward. These routers could be in the same Area or different Areas. Each router only has one path to the other router. There isn't a need to use much bandwidth or CPU in this topology.

Full Mesh Topology

However, bandwidth usage and CPU overhead quickly grow as devices are added to the topology.

Full Mesh Topology

  • Bandwidth Intensive
    • OSPF routers send a lot of messages to neighbors regarding their link state(s). These message are called Link State Advertisements (LSAs). Every router needs to generate and send LSAs to every other router in the topology. After every router updates its routing table...it needs to flood LSAs again to every other router with this update...and so on...
  • CPU Intensive
    • Every router in a link-state protocol calculates a complete map of the network for itself from its Area’s perspective. They do this by comparing received LSAs with its Link State Database (LSDB). Routers run the CPU intensive Dijkstra algorithm to calculate the optimal route for every destination in its routing table.
  • Size of LSDB
    • The Link State Database can get quite large.
  • Issues like a flapping interface can wreak havoc throughout the topology.

To limit the amount of LSAs sent around the network and to limit the use of the Dijkstra algorithm, OSPF uses areas.

Areas

Using Areas greatly simplifies the OSPFv3 domain:

OSPFv3 Areas
Bandwidth and CPU overhead have been greatly reduced. 

Each Area is now concerned only with LSAs within its own area...with the exception of the LSAs generated from the router that borders the other Areas. 

Let's take a closer look at the concept of border routers...

Border Routers

With a border router, routers within an Area don't need to know about every router outside its Area. It can simply use a default/summary route.

There are two types of border routers:

  1. Area Border Router (ABR)
  2. Autonomous System Boundary Router (ASBR)

Area Border Router (ABR)

Inter-area traffic flows through an Area Border Router (ABR). ABRs can connect two or more areas.

Area Border Router (ABR)
Area Border Router (ABR)

Diagram Note: The yellow arrow illustrates traffic flow between the areas, the ABR could decide to send the traffic to either link within the area.

Autonomous System Boundary Router (ASBR)

An Autonomous System Boundary Router (ASBR) connects the OSPF domain to another Autonomous (AS) system. Routes from the other AS are redistributed into the OSPF process.

The ASBR can redistribute the following from the another AS into the OSPF domain:

To continue reading, please login or become a member for full access...