https - hopefully not too stupid a question

[Roland Perry]

In article <87vcipuqo6.fsf at mid.deneb.enyo.de>, Florian Weimer 
<fw at deneb.enyo.de> writes
>>> IPv6 does not mean that the network will provision multiple addresses
>>> to you.  You will still have to pay extra for that.
>>
>> Actually it's exactly the point; over on ORG-discuss an ISP guy who does this says:
>
>[snip]
>
>> …which will give each user's home 2^64 reachable IP addresses.
>
>Not in datacenters, where you'd typically have machines which run data
>centers.  With IPv4, most hosters tell you to configure a /24 netmask
>on your interface.  But this doesn't mean that you'll be able to use
>more than the one IPv4 address the network has handed to you.
>
>With IPv6, there are still technical reasons for limits, although
>addresses aren't scarce: each address requires precious TCAM space,
>and neighbor discovery does not work at scale.

The situation depends on how the "end site" is defined.

RIPE says it's:

    An End Site is defined as an End User (subscriber) who has a business
    or legal relationship (same or associated entities) with a service
    provider that involves:

    that service provider assigning address space to the End User
    that service provider providing transit service for the End User to
     other sites
    that service provider carrying the End User's traffic
    that service provider advertising an aggregate prefix route that
     contains the End User's assignment

So, is the room full of (several customers') racks the end site, or is 
it only the racks for one particular customer.

It's the end-site which gets the /64 (or whatever), the datacentre 
operators gets the /48 (or bigger).

Although the above scenario is a corner case model, because it assumes 
all customers are buying their connectivity from the data centre, 
whereas in practice they may well have their own, or bought-in, carrier 
connectivity.
-- 
Roland Perry