[6bone] pTLA request by NECTEC-TH - review closes 10 March 2003

[Bob Fink]

Philip,

At 01:58 PM 2/27/2003 +1000, Philip Smith wrote:
>Hi Bob,
>
>At 18:40 26/02/2003 -0800, Bob Fink wrote:
>>Before I put a pTLA request out for review I ask the requester why they 
>>aren't getting a production prefix. Generally it boils down to wanting to 
>>try IPv6 services without committing to a production prefix, sometime for 
>>cost reasons, sometimes for organizational reasons.
>
>I can see there being a cost implication for a non-RIR member, but for an 
>existing RIR member? I'm just wondering if the RIRs maybe need to do more 
>publicity to their membership about the availability of production v6 
>space (I somehow doubt this), or, as I just mentioned to Bill, there is 
>something special about 3ffe::/16 which can't be satisfied by production space.

There are differences between the RIRs and what and how they charge.

There isn't anything different about the 3FFE prefix these days (now that 
production prefixes are available) other than they can be gotten for early 
trials and for free. This is not to undercut the RIRs (this has all existed 
for 7 years now, well before the RIRs handed out v6 prefixes) rather to 
make sure there is an early environment for trying out (often called 
testing and experimentation) IPv6. We all too often focus on this being 
free to the pTLA holder, but the greater reality is the no cost method it 
provides to the downstream users of these pTLAs.

That is, there are many thousands of user end-sites, and some intermediate 
networks, that are able to try out IPv6 without having to search out a 
production IPv6 vendor (and there are still very few), make a contract with 
them, and pay real money that is often not available at this stage. A good 
example of this are the various automatic tunnelling sites (like Viagenie 
et al) that serve many many users with no special arrangements or costs 
other than setting up the service for the user. These uses are meritorious 
at this very early stage of IPv6.

When it is reasonable to end the 6bone service is being determined as we 
speak through the 6bone phaseout planning discussions. It does need to end 
sometime, but we want to make sure it doesn't go away until it isn't needed 
during the early deployment stage (that we are at now).


>BTW, I am definitely not suggesting denying NECTEC their request, just 
>curious to understand what technical differences exist between 3ffe::/16 
>and 2001::/16 space which causes them as an APNIC member to not simply get 
>their space from APNIC. APNIC allows people to return production space (as 
>they do for IPv4 space), and production space can be used for trying 
>things out (as we know in IPv4-land).

Can't comment on that.


Thanks,

Bob