[arin-ppml] 2-octet and 4-octet ASNs

Owen DeLong owen at delong.com
Fri Apr 18 20:05:24 EDT 2014

As I understand it, the perceived issue is:

RIRs issuing primarily 4-octet ASNs are running out of 4-octet ASNs and want to get more 4-octet ASNs from IANA without having to first exhaust their 2-octet pools.

Now, for my opinion on the matter:

IMHO, the intent of the current policy was to achieve 4-octet adoption and provide a predictable path to 2-octet ASN exhaustion.

The current circumstance appears to be an unintended consequence of 4-octet adoption exceeding expectations at least in some regions. As such, I suggest the following:

(Hopefully uncontroversial):
1.	Authorize IANA to distribute 4-octet ASN pools to the RIRs without regard to their consumption of 2-octet ASNs

(Probably controversial, but hopefully not too controversial):
2.	Distribute the remaining 2-octet ASNs in the IANA pool to the RIRs on an even basis with any remainder being distributed in descending
	order of number of 4-octet ASNs issued by the RIR in calendar year 2013.

	(IOW, if N%5==3, the 3 RIRs that issued the most 4-octet ASNs would receive 1 each).

However, I believe that both of the above possibilities would require global policy in order for IANA to act on them and I'm not sure such can be achieved prior to being overtaken by events.



More information about the ARIN-PPML mailing list