[IPv6]Re: Analysis of Ungleich ULA Registry

[Nico Schottelius <[e-mail]>]

Hello Mark,

I think I have to change the notion of network size reasoning a bit in
this discussion:

Mark Smith <[e-mail]> writes:
> [... ]
>> 581 Names with >1 Entry
>> 8 Names with >10 Entries
>
> Are you saying that more than 10% of the 4886 unique names have a
> realistic need for between 131 072 /64s (2 x /48s) and 655 360 /64s
> (10 x /48s)?
> [... ]


I don't think that the number of /64's is the main driver ever for
getting a new /48. From the GUA usage we are trained and conditioned
to say that "one /48 equals one location".

I am aware that this discussion is about ULA *not* being in the DFZ,
however I would assume that most people will apply the same logic, one
/48 ULA per site as well. And it makes sense if you have default DFZ
filters in your mind.

If an organisation is using both GUA and ULA (internal), as an
operator only accepting /48's or shorter from other ASNs is just
consistent, even if the ULA originating ASNs are all private and
(mostly?) under our own control.

In a nutshell what I want to say is that it's not about the number of
addresses in a /64 or the number of networks in a /48, because as the
IPv6 community [tm] we have set the notion that a /48 belongs to a site
with all it's advantages and disadvantages.

And people apply it. Thus potentially needing more ULAs.

Best regards,

Nico

-- 
Sustainable and modern Infrastructures by ungleich.ch