[Irtf-rr] IP Address Distribution

Wu Min wumin_noname@hotmail.com
Mon Dec 9 17:00:02 EST 2002


Hi,

Thanx for the links. I find them very useful. I have another question. Is 
there any specific algorithm for aggregating prefixes?



>From: Jaideep Chandrashekar <jaideepc@cs.umn.edu>
>To: Wu Min <wumin_noname@hotmail.com>
>Subject: Re: [Irtf-rr] IP Address Distribution
>Date: Fri, 6 Dec 2002 11:12:12 -0600
>
> > Hi Jaideep,
> >
> > >Hi,
> > >   I'm not really sure what you are asking for.
> > >   Are you looking for geographic allocation of IP blocks ?
> >
> > Yes, precisely and....
> >
> > >   Or are you interested in finding the address space owned by each AS
> > >   in the Internet.
> >
> > I need synthetic one.
> >
> > >   The former is documented in the IANA -- to some extent.
> > >
> > >   For the latter -- it isn't too difficult to find out from the
> > >   route-views BGP table.
> > >   Of course, this data is going to be slightly off because of CIDR
> > >   aggregation -- but I've looked at this in the past -- and the
> > >   aggregation effect is minimal, and should not matter if you want to
> > >   generate synthetic distributions.
> >
> > I have no problem with CIDR aggregation. In fact, I only need to use the
> > addresses to know where to direct specific traffic and all the 
>nitty-gritty
> > details are not required since I do not wish to include end hosts, but 
>just
> > the network core.
>
>
>Just so I understand correctly, you want to generate distributions of
>the IP address owned by an AS, and this is not related to
>geography. If you do need geographic distributions -- there is a
>document on IANA (saw it a while ago) -- that describes the number of
>address blocks assigned to countries (or regions). Note that all of
>them need not be used or even allocated.
>
>For example India is assigned a bunch of class B addresses. If this is
>the sort of information you require -- the best place to look are the
>address registries. ARIN, RIPE, APNIC. I found a website that actually
>lists some of these addresses by country.
>http://www.idefense.com/Intell/CI022702.html
>
>
>On the other hand, if you are interested in a distribution of AS size
>vs. IP address originated, then there are easy ways of getting this
>from the BGP tables.
>
>The following link has a breakdown of prefix ownership by AS.
>http://bgp.potaroo.net/cidr/msprefix-list.html
>
>Or you could generate this directly from the Route-Views BGP table,
>just map each prefix to the origin AS.
>
>
>A long time ago I wondered if there was some sort of pattern in the
>distribution of AS size and Originated Prefixes --- and to the best of
>my knowledge there isn't any.
>To be clearer -- there is no indication that if size(A) > size(B),
>then A originates more address blocks that B. Here size(A) is the
>degree of A.
>
>
> > >   From what I saw a while ago -- there is very little correlation
> > >   between AS degree and IP address space.
> >
> > You reckon IP address space can be distributed randomly?
>
>
>I hope this helps.
>
>
>
>
>-jaideep


_________________________________________________________________
MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*. 
http://join.msn.com/?page=features/virus




More information about the irtf-rr mailing list