North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: Top-50 Report

  • From: Brett D. Watson
  • Date: Sat Sep 14 14:15:14 1996

> 
>  Paul A Vixie <[email protected]> writes:
>   * > Remember, this just looks at the classful routes and tries to cidrize
>   * > them at the origin AS level. It assumes no holes either so it's quite
>   * > possible to get a lot more back. Yes there are 31K classful routes. That'
>   * s
>   * > either a lot of legacy or not a lot of aggregation.
>   * > 			--Tony
>   * 
>   * Given that the routing table stood at 28000 when CIDRD was killed off, I'd
>   * say that there have either been some recent classful allocations, or a lot
>   * of ancient networks suddenly getting advertised.
> 
> Or a lot less care and letting classful size announcments leak out and
> not aggregating to the allocation folks were given. To highlight I pick
> the last 50 classful entries I see in the 193, 198 and 206 range just
> to give a random sample and to illustrate.

  this is what i was getting at in my note a week ago or so.  it 
appears that many more networks are being announced by 2+ as's.  so 
i'm assuming we're sort of doomed to poor aggregation as more and 
more people dual-home to providers.

  and i still see some pretty poor aggregation (like /22's and /24's) 
in some recent allocations.  

-brett

- - - - - - - - - - - - - - - - -