[c-nsp] BGP prefix announcement question..

[John van Oppen]

That is defiantly the way to go...  Sadly lots of small networks still
use prefix lists for everything and leak interesting routes every time a
customer turns off their BGP session.

John 

-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Pete Templin
Sent: Wednesday, June 25, 2008 8:14 AM
To: Drew Weaver
Cc: cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] BGP prefix announcement question..

Drew Weaver wrote:
> Hi there, there appear to be a few different ways (I'd almost say
> "many") to announce/filter BGP prefix announcements between neighbors
> these days (I'm speaking specifically about Internet edge neighbors
> here i.e. company to internet transit provider), I am trying to see
> if there has ever been defined a 'standard' way to do it which makes
> things like employing Blackhole communities (on the side of the
> corporation) a little less painful than cutting your own teeth with a
> spoon.

Every BGP prefix in our network gets tagged with at least one community,

that of a "magic code", upon origination/learning/injection/whatevah. 
It's basically of the form <ourAS>:ABCDE, where A indicates the type of 
route (customer, ours, public peer, private peer, paid transit) and BC 
indicates the POP of first contact.  We then filter based on the :A code

before propagating to upstream providers.

Tedious to author, but phenomenally useful once in place.

pt
_______________________________________________
cisco-nsp mailing list  cisco-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/