The Cidr Report

Christopher L. Morrow christopher.morrow at mci.com
Sat Nov 13 03:31:26 UTC 2004



On Fri, 12 Nov 2004, Randy Bush wrote:

> >>> ASnum    NetsNow NetsAggr  NetGain   % Gain   Description
> >>>
> >>> AS18566      751        6      745    99.2%   CVAD Covad Communications
> >>> AS4134       825      178      647    78.4%   CHINANET-BACKBONE
> >>>                                                No.31,Jin-rong Street
> >>> AS4323       794      223      571    71.9%   TWTC Time Warner Telecom
> >>> AS6197       814      430      384    47.2%   BNS-14 BellSouth Network
> >>>                                                Solutions, Inc
> >>> AS22773      401       17      384    95.8%   CXA Cox Communications Inc.
> >>> AS27364      413       45      368    89.1%   ARMC Armstrong Cable Services
> >>> AS701       1230      884      346    28.1%   UU UUNET Technologies, Inc.
> >>> AS22909      412       81      331    80.3%   CMCS Comcast Cable
> >>>                                                Communications, Inc.
> >> e.g. is AS18566 the origin AS for 751 prefixes that could be
> >> collapsed to 6?
> >
> > not to justify the expense, but perhaps covad is renumbering from one
> > block to another? Looking at their advertisments I see lots of /23 or /24
> > blocks inside their larger covering routes... So either they deaggregated
> > to renumber more gracefully, or they forgot their prefix-list outbound to
> > williams and exodus ?
> >
> > perhaps covad can explain? or silently cover up the 'mistake' (which is
> > acceptable as well...)
>
> it's not just covad.  they're just such an egregious case among
> many socially and technically irresponsible polluters.

eh, since I singled out covad: (and I feel bad for it now)
what about for COX? what about for UU (doh, thats me...or our tac or
something, I'll look/ask), what about Comcast? and TWTC? ArmStrong?

Of these listed 4 are cable companies, is there something in the cable
modem networking that requires deaggregated routes beyond their borders?
Is the problem that they might have seperate 'networks' for their regional
parts and leak more specifics for these parts along with 'backup' routes
via aggregates?

-Chris



More information about the NANOG mailing list