Re: [Idr] Update on advancement of rfc5575bis draft, implementations

[John Scudder <[e-mail]>]

On May 19, 2019, at 2:15 PM, Christoph Loibl <[e-mail]> wrote:
> However, I do not 100% understand the correct process regarding the traffic-rate-packets code-point. If we want to assign it from the IETF review code space do we need to ship the RFC first (if this is the case, I sense a deadlock condition)?

IETF Review code points are indeed ordinarily assigned when the RFC is shipped. You’re correct this doesn’t work right with our IDR procedures. Jeff has proposed that in this case we simply move ahead with shipping the RFC regardless, if we do so it would cause the code point to be allocated with no further ado.

Otherwise, the workaround is the procedures from RFC 7120, "Early IANA Allocation of Standards Track Code Points”. You can refer to the RFC for the full details of how it works, but the summary of how we do it in IDR is:

1. Someone (typically but not required to be an author) requests early allocation of a code point.
2. Assuming the chairs don’t have some objection (there would be none in this case), the chairs offer the WG a chance to object (typically a two-week consensus call).
3. Assuming there’s support (sometimes in the form of “no objections”) the chairs proceed to request the early allocation. 
4. Before the allocation can be done, the Area Director also has to approve.

As you can see, the process can be expected to take +/- 3 weeks or so, once a request is made.

The First Come, First Served (FCFS) procedure (and much more) is in RFC 8126, "Guidelines for Writing an IANA Considerations Section in RFCs”. You can also see https://www.iana.org/help/protocol-registration for a concise version. The summary is:

1. Someone (typically but not required to be an author) requests allocation of a code point using the IANA form https://www.iana.org/form/protocol-assignment
2. There is no step 2. That’s all. IANA typically turns the request around in one working day or so.

As I mentioned earlier the registry in question has FCFS and IETF Review ranges, so any of the three solutions above can work (but for the first solution to work, we’d have to waive the normal WG process).

Regards,

—John