Deaggregation Disease

Fri Jul 21 14:48:10 UTC 2006

On 21-Jul-2006, at 09:17, Rob Evans wrote:

> There seem to be a whole load of ASNs that have deaggregated.  AS5416,
> AS5639, AS6140, AS9121, AS13049, AS16130, AS17849,  AS18049 (that's as
> far as I got before getting bored).  Some of these are advertising the
> covering prefix too, so they're certainly aware of how to aggregate.

Sometimes this is done intentionally -- the long-prefix (covered)  
prefixes might be TE routes designed to draw traffic to particular  
sinks through specific external providers.

People have been known to stamp NO_EXPORT on those and get some  
measure of TE without polluting the global table, but if the AS whose  
exit you're trying to influence isn't adjacent that doesn't work.

As it happens, Tony Li, Rex Fernando and I wrote up a proposal for a  
new attribute which might help in some of these situations. (It's a  
crude mechanism, but not as crude as NO_EXPORT).

    http://www.ietf.org/internet-drafts/draft-ietf-idr-as- 
pathlimit-02.txt

Under that proposal you could stamp an AS_PATHLIMIT=2 or  
AS_PATHLIMIT=3 on TE routes and have them automatically dropped by  
routers when the AS_PATH length exceeded 2 or 3. For some people this  
would work (for others, for whom 90% of the Internet is less than 2  
or 3 hops away it wouldn't do much).

It would help immensely with getting that document published if  
people could read that draft, and let me know if it looks like  
something they would implement if it was implemented. Private mail  
would be great.

Joe