I attended the ARIN 36 meeting that was held in Montreal in October, immediately following NANOG 65.
The logistics of the meeting were up to ARINâ€™s high standards: the agenda, the discussion guides, the information provided to attendees was well executed, as usual.
The agenda included a number of policy discussions, some reports and candidate speeches for election to the various open slots in the Address Council, the Advisory Council and the Board of Trustees.
In terms of the policy discussions, the group considered out of region use of addresses, and came to what I personally feel is a reasonable conclusion on this particular matter, namely that they can accommodate this with little fuss. Given that ARINâ€™s IPv4 address pool is now entirely committed, and IPv6 is simply not a problem it seems to be more of a formality than a large scale change of position, but it is a stark contrast to the recent LACNIC debate along the same lines which concluded with what I understand to be an opposite outcome. My understanding of the genesis of the RIRs was to better serve local communities with their IP address needs, not to erect fences and define quarantine zones, but of course scarcity tends to push folk to unusual outcomes. Also of note in the ARIN policy debates is the move away from â€œdemonstrated needâ€ when processing market-based address transfers. Itâ€™s been a piecemeal erosion of the concept rather than abandoning the concept completely in a single stroke of policy, but it seems to me that this process is now underway following the exhaustion of the ARIN free IPv4 address pool. There is a certain irony in a developing situation where it may well be that the only RIR still applying some form of evaluation process of â€œneedâ€ as a precondition (or impediment) of registering a transfer is APNIC, and as I recall the AP community adopted this only because they felt that it was a pragmatic move to align their policies to those of ARIN some years back!
Aside from these matters was little of major substance in the policy items considered at this meeting. IPv4 allocations have all but finished for ARIN and they have a waiting list for residual returns. They have largely stopped tinkering with IPv6 allocation policies, and thatâ€™s probably a good thing.
However, Iâ€™m going to have to be somewhat critical of this meeting in terms of its objectives and style. It appears that RIPE and ARIN have gone down paths with their meetings that are very different. The RIPE community has all but pushed the policy process onto the mailing list, and their meetings do not dive at any depth into debating policies. The meeting style for the RIPE community is largely themed around information sharing and collaboration, and this appears to have been a highly successful decision for them. On the other hand, the ARIN meetings have headed further down a path of being dominated by debates about policies, with some statutory reporting tacked on. This policy debate at the ARIN meetings is now concentrated with just a few folk, and it appeared for me that there were times that somehow I had been caught up in an internal meeting of the ARIN Advisory Council as a passive observer. Maybe the changing nature of the environment, particularly with the exhaustion of ARIN’s IPv4 free pool, represents a good time for ARIN to re-examine the intent of these meetings. It could be possible for ARIN to take a leaf from the RIPE community meeting approach and introduce a little more in the area of information, analysis and collaboration in their face to face meetings, and use the mailing list more strenuously as the major venue for their policy development process.