[lacnog] Subasignación de prefijos a otro ASN y el tema con los RoA

Fernando Frediani fhfrediani en gmail.com
Vie Feb 4 18:04:41 -03 2022


On 04/02/2022 17:16, Mike Burns wrote:
> <clip>
> As you say, it’s a simple matter of comparing ASN advertisements to the
> block ownership being advertised, and then check to see if the block being
> advertised by the foreign ASN is legacy. Pretty simple and I think that may
> convince you.
Not just that.
If the block under these conditions is registered in LACNIC, AFRINIC or 
APNIC it may be worth to investigate under which conditions that is 
happening in order to find out if it is being leased and let the RIR 
know or simply used by the resource holder under the allowed conditions.
>
> You may also read the threads from the ARIN ppml I posted earlier to hear
> feedback on this precise issue from many others, not just me. Kind of like
> the due diligence you have advised, use multiple sources.
>
> Of course we know that many cloud providers advertise blocks belonging to
> their clients, not themselves, and they advertise them under the cloud
> provider's ASN.  AWS, Cogent, Oracle, Vultr, etc. Not sure how you would
> differentiate those from leases, but with enough investigation you are sure
> to find smaller, non-cloud provider ASNs advertising space belonging to
> others.
>
> If you would agree to non-disclosure I will share an ARIN non-legacy block
> being leased by a Latin American client off list as I believe I can acquire
> their permission. They know they are breaking no rules leasing ARIN RSA
> space.
As far as I now, unfortunately in ARIN there is not currently impediment 
to lease ARIN non-legacy blocks.
>
> You still say non-legacy space is not safely leasable, but you can't point
> to any policy or RSA language that prevents RSA addresses from being leased.
> Only in LACNIC and AFRINIC is there a risk, not that I've ever heard of
> LACNIC addresses revoked for usage differing from the original
> justification. Maybe you have?
>
> If what you say is true (that non-legacy space can't be leased), where is it
> in the RSA or the NRPM of ARIN, RIPE, or APNIC?
>
> Looking forward to your pet-project results.
>
> Regards,
> Mike
>
> -----Original Message-----
> From: LACNOG <lacnog-bounces en lacnic.net> On Behalf Of Rubens Kuhl
> Sent: Friday, February 04, 2022 1:16 PM
> To: Latin America and Caribbean Region Network Operators Group
> <lacnog en lacnic.net>
> Subject: Re: [lacnog]
>         Subasignación de prefijos a
> otro ASN y el tema con los RoA
>
> On Fri, Feb 4, 2022 at 2:50 PM Mike Burns <mike en iptrading.com> wrote:
>> Hi Rubens,
>>
>> I'm going to ignore the childish ad hominem.
> Not ad hominem, apply to all brokers, not just you. The same with the pet
> project I will now start to identify all RIR leases that are not kosher, not
> just your clients.
>
>
>> My facts stand, yours were proven incorrect.
>> You were the one saying only legacy addresses can be leased. Still
>> standing by that one?
> Still standing by that one, and they were not proven incorrect by anything
> in this thread.
> You not published the lease list also didn't help make a point either way,
> but that will be solved soon enough by some processing of RIR delegated
> lists and DFZ BGP announcements.
>
>
>> You won't point to any policy or RSA that prevents leasing and you
>> ignore the many lease providers operating today.
> You assume those lease providers are operating correctly, something I don't.
> I also don't assume they are operating incorrectly, and every step of the
> way I mentioned that each allocation is unique in its ability to be a part
> of a lease or not.
>
>
>> My facts are:
>>   that only LACNIC and AFRINIC registered blocks pose any risk at all
>> due to RSA language
> It's good that someone reckons this, because sometimes (not you) I read
> otherwiser.
>
>>   non-legacy space is leasable safely
> That's where we differ.
>
>>   leasing with ROA is being done today and I  am the only one who
>> answered the OP correctly on his question.
> I also mentioned that about ROA, so you keep getting facts wrong even when
> they are fresh.
>
>
> Rubens
> _______________________________________________
> LACNOG mailing list
> LACNOG en lacnic.net
> https://mail.lacnic.net/mailman/listinfo/lacnog
> Cancelar suscripcion: https://mail.lacnic.net/mailman/options/lacnog
>
> _______________________________________________
> LACNOG mailing list
> LACNOG en lacnic.net
> https://mail.lacnic.net/mailman/listinfo/lacnog
> Cancelar suscripcion: https://mail.lacnic.net/mailman/options/lacnog


Más información sobre la lista de distribución LACNOG