Joel, Glen, Le 24/09/2011 03:18, Joel Maslak a écrit :
Protection against learning a bad default route through whatever routing protocol they are learning, since these two routes would be more specific than any typical default route. They probably got burned learning a default route.
Having a default route, or rather having a route to every possible adresses, is required when you expunge your routing tables of some prefixes yet you still wish to contact them relying on the next-hop's table. Simple application is to filter incoming routes longer than /20 or /21 to free up some memory on your routers (reducing the global table from 377k to less than 100k routes is a nice perspective ;) ) But a default route is an obvious move and could easily be leeked by an upstream, yet replacing yours if not properly filtered. So, using more precise routes (/1s to /8s) helps avoiding these risks and yet lets you roughly balance load to several gateways. -- Jérôme Nicolle