Routing

i do not understand the usage of aggregate route.

form the junos traninig material, the example is like this.

R1 has interface 1 connect to network 172.29.0.0/24; interface 2 connect to 172.29.1.0/24, interface 3 connect to 172.29.2.0/24.

Then, there is a aggregate route 172.29.0.0/22 next-hop reject.

so, any destination in 172.29.0.0/22, but, not in those 3 networks will drop.

my Question is ....... why bother?  the is no network in 172.29.0.0/22 other than those 3 networks.  if some one send a destination to 172.29.0.0/22 (but, not in those 3 networks), even there is no aggregate route, it will drop anyway, since destination not found. 

We typically use aggregate routes to decrease the number of routes you need to propogate around a network. 

For example, with service providers we have large allocations like /19 or /20 that we send to our peering partners.

Inside our network not all of these are in use yet but small sections /30 /29 etc are assigned daily to new turn ups.

Aggregate routes provide a simple way to send the entire block up knowing only some is in use and keeping the overall internet table smaller.  Also in this case anything smaller than a /24 cannot be passed to the peer either.

Enterprise networks use a similar method when they allocate larger blocks of say /23 to /16 space to a logcial section of the network.  Say the enterprise has two data centers, 2 large sites and 2 WAN connection methods.  They assign a large block to each of the 6 general areas.  then at the connection points between the sections they only need to push the aggregate route for the other areas into each other.  the routing table becomes very small a default route plus the 5 aggregates for the other areas and the individual areas internal table.

There is also less route churn then using these methods dropped links affecting small subnets inside the aggregate only affect routers in the local area and don't have to be spread around the entire network.