Prefix hijacking by Michael Lindsay via Internap

David Conrad drc at virtualized.org
Sun Aug 21 19:26:08 UTC 2011


Jimmy,

On Aug 21, 2011, at 8:15 AM, Jimmy Hess wrote:
> The system is this way BY DESIGN,  and any other method would concentrate power
> which would be  detrimental to the internet  and counter to its
> open/consensus driven nature.

See recent discussions in RIPEland regarding BGPSEC+RPKI regarding policy 2008-08.

> The RIRs' job is to provide unique registrations, nothing else.
> And registry fees are for recovering costs necessary to provide the
> service and to maintain addressing policy.

Googling "<RIR> budget 2011" gives (all in US$, using today's Euro and Aus$ conversion rates):

AfriNIC: 2011 Expenses: $2,832,614 (http://www.afrinic.org/corporate/Budget2011.pdf)
APNIC: 2011 Expenses: $14,815,906 (http://meetings.apnic.net/__data/assets/pdf_file/0005/31478/treasurers_report.pdf) 
ARIN: 2011 Expenses: $16,412,160 (https://www.arin.net/about_us/corp_docs/budget.html)
RIPE: 2011 Expenses: $26,313,894 (http://www.ripe.net/ripe/docs/ripe-507, using today's exchange rate)

Can't seem to find a 2011 (or anything more recent than 2005) budget for LACNIC.

BTW, since you made passing reference to them and just FYA:

ICANN: Expenses: $61,164,000 (http://www.icann.org/en/financials/adopted-opplan-budget-fy12-09aug11-en.pdf)

So if LACNIC's budget is more than $790K, the RIRs combined budget for "providing unique registrations, nothing else" for IP addresses will be more than ICANN's budget for coordinating all unique Internet identifiers including addresses, IETF protocol parameters, and domains (and "promoting competition" in the latter) and dealing with a much larger community (at least as measured by meeting attendance).

> _Enforcement_  of   RIR  allocations is by network operators   refusing to
> originate or propagate announcements  by organizations unauthorized by
> the registered resource holder.

Very true. Of course, defining "registered resource holder" and finding that information will likely get a bit more complicated in the future...

Regards,
-drc





More information about the NANOG mailing list