>> /48 was intended to be the *minimum* allocation to a subscriber's site, not
>> the *maximum*. Those exceptional subscribers for whom a /48 is too small
>> are free to request larger blocks from their ISPs.
>But that's where the /35 "pressure" applies because if (say) UKERNA wanted
>to offer a /40 to each University it would at present be looking at being
>able to connect just 32 Universities. By 2001 there will be over 800
>University and further education colleges online under the UK JANET network.
While I understand the problem you have, I would like t know why
you'd like to offer /40 to every universities. Are there any reason
not to do /48 for universities, or couple of /48s?
We (WIDE) offer /48 for leaf sites (an organization which has no
plan to have sub-organization, i.e. universities and companies,
not an ISP), and /40 to non-leaf downstream sites (an organization
which has plan t have sub-organization, i.e. an ISP).
Once a leaf site use up /48, they'd be able to get another /48 from us.
Until now we saw no problem with it.
"/48 allocation to everyone" is just fine for me (or us).
We cannot predict the future growth of leaf site.
We don't want to, we don't need to. SOHO can evolve into
multinational company.
We have more important problem to tackle. The problem lies elsewhere,
I believe. Once our network grows into /29, and to /16, we'd need
to carry 2^(48-29) or 2^(48-16) routes in a TLA. It's like having
the whole IPv4 space in one ISP. We need to have a very careful plan
on inner-TLA topology, so that we can aggregate internal routes nicely.
We may need more experiences/standards in hierarchical interior routing
operation.
itojun
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page: http://playground.sun.com/ipng
FTP archive: ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------
