FW: Net-block issues

Roeland Meyer rmeyer at mhsc.com
Wed Aug 15 15:27:40 UTC 2001




|> -----Original Message-----
|> From: Roeland Meyer [mailto:rmeyer at mhsc.com]
|> Sent: Wednesday, August 15, 2001 8:24 AM
|> To: 'Elisabeth Porteneuve'; ga at dnso.org; kent at songbird.com
|> Subject: Net-block issues
|> 
|> 
|> 
|> |> From: Elisabeth Porteneuve 
|> [mailto:Elisabeth.Porteneuve at cetp.ipsl.fr]
|> |> Sent: Wednesday, August 15, 2001 1:58 AM
|> 
|> |>     The rivers of comments has been sent out about domain
|> |>     names, there is a dangereous silence about IP numbers.
|> 
|> It is with wry amusement that I have been following the MAPs 
|> debate, on
|> NANOG, recently. Then you make this comment, it couldn't 
|> dove-tail more
|> beautifully. Those that do not know what MAPs is, please review
|> http://www.mail-abuse.org/. 
|> 
|> The fundimental function is to create blacklists used for 
|> various filtering
|> of IP blocks. While this sounds innocuous on its face, MAPS 
|> has recently
|> announced intention of charging for their service, in order 
|> to raise money
|> for legal fees (they are under legal attack). Until now, 
|> Paul Vixie has been
|> the main funder of MAPs.
|> 
|> The problem is that MAPS was distributing the lists and 
|> there were many
|> local copies, within places like EarthLink and AOL. As long 
|> as those copies
|> are maintained by MAPS, this is not a serious problem. They have an
|> efficient and timely distribution mechanism. The result of 
|> an net-block
|> being listed in MAPS is that entire chunks of the Internet 
|> cannot reach that
|> net-block. This is done at the provider level. The effects 
|> are global in
|> that they over-ride ARIN, RIPE, and APNIC.
|> 
|> The problem arises when there are stale entries in the 
|> black-list. In fact,
|> the danger has always been, the issue of stale entries. The 
|> effect, given
|> MAPS market-share, is that a net-block can be issued, to a 
|> new business,
|> that may be perfectly useless and neither the ISP or the 
|> business knows
|> about it until it is too late. I might point out that a 
|> similar problem can
|> occur with DNs, so it is not entirely an PSO issue. There 
|> are now stale
|> entries, in abundance, because many providers are still 
|> evaluating the new
|> cost issues. Meanwhile, they have disconnected from the 
|> distro system and
|> aren't having their black-lists updated. Ergo, they have 
|> stale entries.
|> 
|> What this, in effect, does is to over-ride various registry 
|> policies. Names
|> and net-blocks that are issued may not, in fact, be issued. 
|> Names are not a
|> serious cost issue and they can, with some logistical difficulty, be
|> re-issued. However, net-blocks cost over $1US/month (at the 
|> ISP level), or
|> $2500/year from ARIN (the US registry), and they are not all 
|> that plentiful.
|> 
|> The fundimental process disconnect here is that, IMHO, the various
|> registries should be performing the MAPS function as part of 
|> their policy
|> enforcement mechanism. This is not written into any of the
|> registrar/registry agreements.
|> 
|> If anything argues for a centralised systems approach, the MAPS
|> functionality does. IMHO, this makes it an ICANN issue. Yes, 
|> this also
|> politicizes it somewhat. No, some NANOG denizens won't like 
|> it and that is
|> guaranteed.
|> 
|> --
|> R O E L A N D  M J  M E Y E R
|> Managing Director
|> Morgan Hill Software Company
|> t:01 925 373 3954
|> c:01 925 352 3615
|> f:01 925 373 9781 
|> 



More information about the NANOG mailing list