[dp-tf] Re: 91.198.71.0 - 91.198.71.255

[Richard Cox cio@localhost]

(Copied to APWG-Chairs and DP-TF in view of the policy implications)

On 02/11/2007 16:36:35, Alex Le Heux alexlh@localhost wrote:

> As it turns out, we had already noticed the change in registration
> data a few weeks ago and are currently in contact with the LIR.
> We have established procedures that are designed to track changes
> like this, and this case is already under investigation.

There appear to be seven other ranges and ASNs, similarly acquired.

	193.33.128.0/23  AS42672	194.110.69.0/24  AS42811
	91.193.40.0/22   AS42662	91.193.56.0/22   AS42672
	91.194.140.0/23  AS43188	91.195.116.0/23  AS43702
	91.196.232.0/22  AS43259	91.198.71.0/24   AS43603

> We have audited the request and our policies and procedures
> were correctly applied when this range was assigned.

That's obviously very worrying - because it implies that any similar
new application now would also be granted, and the claims that RIPE
still has three years to go before IP4 exhaustion would then be seen
as having been highly optimistic!

I note your earlier comment about the data having been recently changed:
so are you saying that the original application was valid just because
it had an address in the RIPE region - even though that address may have
been meaningless?  If so, are you able to remind us what that address
originally was (as at the time it would have been "public" data)?

As you may know, the entity believed to be using these IP assignments is
the notorious "Russian Business Network" about which much has recently
been written: and http://en.wikipedia.org/wiki/Russian_Business_Network
provides a convenient index to the more significant of those articles.

The RIPE NCC cannot, of course, be concerned with any criminal activity
that uses IP addresses it assigns, but I believe the RIPE NCC does need
to be concerned about the obtaining by dishonest means of resources that
are owned by the community and entrusted to the stewardship of RIPE NCC.

During RIPE 55 I mentioned the earlier cases of shell companies set up
apparently by a "Boris Mizhen" to apply for IP resources for spamming
purposes (ie to avoid filters) and I understand that the LIR handling
those applications (Merezha) has previously been linked to questionable
assignments.  For the record, the IP ranges and ASNs involved with that
series of incidents, were as follows:

	91.193.152.0/22  AS42719	91.193.192.0/22  AS42719
	91.193.216.0/22  AS42719	91.193.88.0/22   AS42719
	91.200.124.0/22  AS42719	91.200.132.0/22  AS42719
	91.200.164.0/22  AS42719	91.200.176.0/22  AS42719
	91.200.56.0/22   AS43791	91.200.60.0/22   AS43791
	91.200.72.0/22   AS43799	91.200.80.0/22   AS43799

I (and others) mentioned at the Address Policy WG the probability of an
imminent IPv4 landgrab - these two recent incidents seem to suggest that
it has started.  How well can RIPE policies stand up to such deliberate
and abusive attacks?  In particular, how protected would the community
be against an LIR that intentionally submits applications which rely on
data the LIR knows is bogus?  The LIR is not identified on the WHOIS
output - but if the only policing of the application is done by the LIRs
(as I'm told that RIPE hostmasters are not allowed to question details
supporting an application for resources) perhaps the identity of the LIR
should be displayed against the IP range, so that patterns of dishonesty
can quickly become visible?

> We do sometimes assign resources to organisations for use outside our
> region, although it rarely happens, and such requests are handled very
> carefully as they are rather unusual.

I'd welcome some clarification on the policies involved there: is that
just to entities ("organisations") located, at least nominally, within
the RIPE service region: with operations outside that region - or would
entities/organisations within other regions qualify?  I ask because the
other questionable incident that came to our attention recently was the
use of 85.255.112.0/20 in California USA, when it was shown in the RIPE
database as unassigned space: and we discovered that an assignment was,
mysteriously, made of that space by the RIPE Hostmaster to "Inhoster" on
the very day I pointed out the inappropriate use, and that assignment
has also recently had its data changed: showing it as now being assigned
to a "UkrTeleGroup".  But still it is used - solely - in California USA.

> It is a normal occurrence that a request is denied by one of the RIRs.
> If the reason for the denial is related to the location of the network,
> the end-user is then referred to the correct RIR.  This is normally
> nothing to be suspicious about as there are many legitimate
> organisations that have operations in multiple RIR regions.

Given the (apparent) disparity in policy implementation between RIPE
and the other RIRs - most of whose hostmasters are empowered to (and
indeed do) check for and reject abusive applications it seems RIPE
may soon be regularly targeted by organisations worldwide who cannot,
for whatever reason, get resource assignments from their local RIR.

> Could you tell us a little about the circumstances that brought
> this range to your attention? Any information might help us with
> our investigation.

We have been tracking the "Russian Business Network" and also "Boris
Mizhen" for some time now, and we have been monitoring routing and
other changes which tell us their traffic has moved to new routes.
It is, however, unlikely that their actual location has changed, as
traceroutes to IP addresses in the new ranges are indicative of the
use of a "traceroute simulator" rather than of a real network path.

Best regards

-- 
   Richard D G Cox cio@localhost
   CIO, The Spamhaus Project
   http://www.spamhaus.org