Le 28/03/2013 19:24, Michael Richardson a écrit :
"Alexandru" == Alexandru Petrescu
<alexandru.petre...@gmail.com> writes:
Alexandru> 2 - the prefixes obtained from Registries, or from ISP
(which one Alexandru> should I try first?) may come with a price tag.
The more vehicles, Alexandru> the pricier the allocation.
Depending upon the way in which the VIN->subnet ID is formed, it
might be that each manufacturer should actually go and acquire a /28
or something from their local RIR. That would be one way.
Yes it could be one way to be considered.
At that consideration, one should think whether 38bits is enough to put
more than one /64 in each other vehicle of that manufacturer, for that
length of time.
The RIRs might find that they want to adjust their policies and talk
to IANA, allocate this out of a Non-Connected Network space.
Ok.
That would be, from the IETF point of view, the easiest way to do
this. Creating the subnetID could in fact be manufacturer
proprietary, and the document would just be a BCP on how not to do it
wrong.
It would likely be, from the RIR and automotive industry, a complex
way, because I think neither RIR community nor automative community
knows or understands each other.
I agree. It looks procedural and long. If this was discussed elsewhere
previously (like IANA/RIR lists?) I am interested to look.
A different way would be for the IETF to instruct IANA to create a
NCN space, and designate the VIN-mapping as a part of it. This would
be harder for the IETF to do, but would I think, lead to a better
world.
NCN? Non-Connected Networks? That is another name for ULA space no?
Alexandru> 3 - prefixes which are provider-assigned and/or provider
independent Alexandru> may introduce routing churn in the core of the
Internet - if these Alexandru> prefixes are numerous.
you have confused globally routable with globally unique. These
prefixes are simply *NOT* going to be announced in the core of the
Internet.
Well, confusion may have crept in, sorry if so.
I understand ULA space is not to be announced in the core of the Internet.
I understand PA Provider-Assigned addresses are to be announced in the
core of the Internet.
I understand PI Provider-Independent prefixes are to be announced at
several places, but still restricted (not 'anywhere'), yet be reachable
in the Internet. (these restrictions are clear at RIR; e.g. it can only
be /48 not shorter, unless exceptions to be documented; and more other
clear restrictions).
If they *are*, it will be a manufacturer doing so, and and using one
of many tunneling technologies to get the packets to the right place
(whether it be IPsec, IPIP, MobileIPV6, GRE, or IP-over-Acoutistic
Carrier).
Hm...
Alex
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
