IP Routing: Aggregation Concepts and Techniques

  • Purpose of route aggregation
  • Scalability and fault isolation
  • How to Aggregate

What is the purpose of route aggregation? We aggregate routes to decrease the sizes of our routing tables, hide information, and decrease convergence times. The boundary device – ASBR / ABR / whatever – will use a route summarization technique to aggregate prefixes towards the direction of advertisement. For example, if you own the 10.0.0.0/8 network, and you’re advertising this network to a customer, you may wish to simply advertise the classful network rather than advertising more specific prefixes (like 10.1.1.0/24, etc.) even if you have your network subnetted. By only advertising the aggregate, the size of your customers routing tables is minimized, which decreases the amount of calculation needed during reconvergence and hides information about how your AS is subnetted. The advertisement of a default route towards the access layer is also an example of aggregation. How does one accomplish the task of aggregation with various routing protocols?

  • BGP: the use of the aggregate-address <NET> <MASK> <summary-only> command
    • at least one more-specific address in that NET need to be present in the IGP and either redistributed into BGP or present in a network statement
    • you can use a route to Null0 to create this IGP route
  • OSPF: stub and the area <Area #> range <NET> <MASK> command
    • best practice is to create a route to Null0 to avoid routing loops
    • you may also need to look at no rfc1583 compatible command
  • EIGRP: stub and the ip summary-address eigrp <AS> <NET> <MASK> interface command
    • Remember that when you do this, a route to Null0 is automatically created
    • EIGRP also automatically summarizes, so you may want to disable this with the no auto-summary command

References: OSPF Design Guide EIGRP Design Guide BGP Case Studies

2 thoughts on “IP Routing: Aggregation Concepts and Techniques”

Comments are closed.