Summary

  • ARIN-2025-7 makes the single-site /48 starting point explicit and retains the 75% next-larger-nibble rule for larger multi-site allocations.
  • The bands are 2-12 sites /44, 13-192 /40, 193-3,072 /36, and 3,073-49,152 /32.
  • The proposal is a Recommended Draft under discussion, not Board-adopted or implemented; its effect on applicants and historical requests remains unknown.

ARIN’s Recommended Draft Policy ARIN-2025-7, titled Make Policy in 6.5.8.2 Match the Examples, addresses a mismatch between wording and examples in the IPv6 allocation rule. The current wording says that a qualifying organisation may receive an initial /48, then applies a 75% threshold to the number of /48s in a prefix. Read literally, one site accounts for more than 75% of the single /48 contained in a /48. That arithmetic conflicts with the examples.

The proposed wording separates the cases. One site that meets the initial criterion receives an initial /48. The 75% next-larger-nibble rule applies to larger multi-site allocations, where the justified site count determines the size. The proposal therefore clarifies the calculation; it does not change eligibility criteria.

The complete bands must remain visible: one site receives /48; 2 through 12 justified sites receive /44; 13 through 192 receive /40; 193 through 3,072 receive /36; and 3,073 through 49,152 receive /32. Both endpoints matter at each boundary. The draft also changes assignment to allocation throughout section 6.5.8.2 because ARIN staff say assignments are no longer issued under this procedure.

The Advisory Council advanced the proposal to Recommended Draft Policy on 18 June 2026. ARIN announced that action on the PPML on 24 June 2026. The current policy text is dated 20 May 2026, and the reporting freeze is 5 September 2026. At that freeze, the proposal remains under discussion: it is neither Board-adopted nor implemented.

The draft labels implementation Immediate. Separately, ARIN staff estimate three months for staff training and updates to public documentation. The Advisory Council assesses that the clarification matches current registry procedures and introduces no operational change. Staff assess that the text is implementable as written, has no impact on registry operations and services, and raises no material legal issue. These are attributed source assessments, not independently proven outcomes.

BTW’s boundary-case test is the useful operational lens. Review the transitions at 1/2, 12/13, 192/193 and 3,072/3,073, and confirm that each justified site count maps to one predictable nibble-aligned result. The available sources do not quantify how many historical or future requests fall on those boundaries, whether an applicant received a different size because of the wording, or what later community or Board action will do.

Publication note: Published is the scheduled date, 2026-09-05. The actual server publication timestamp remains unset until the production publish action.