[10] in SIPB IPv6
Re: quick status; perl hack needed..
daemon@ATHENA.MIT.EDU (Bill Sommerfeld)
Thu Oct 7 11:52:40 1999
Message-Id: <199910071523.PAA10659@orchard.arlington.ma.us>
To: Marc Horowitz <marc@MIT.EDU>
Cc: sipbv6@MIT.EDU
In-Reply-To: Message from Marc Horowitz <marc@MIT.EDU>
of "07 Oct 1999 01:20:30 EDT." <t53k8oz6929.fsf@horowitz.ne.mediaone.net>
Date: Thu, 07 Oct 1999 11:23:41 -0400
From: Bill Sommerfeld <sommerfeld@orchard.arlington.ma.us>
> Bill Sommerfeld <sommerfeld@orchard.arlington.ma.us> writes:
>
> >> - zero, one, or many ipv6 prefixes (address and prefixlength pairs)
>
> This implies you can route any downstream IP address.
Not necessarily; the prefixes almost certainly lie within our
allocation.
Also, it's conceivable that we could add other local shortcut routes
not within our allocation as a favor to someone on the other end of a
tunnel (e.g., for freenet6 addresses) but not advertise those prefixes
upward into the core of the 6bone.
> While I believe you could probably inject anything you want into the
> routing tables, aren't we supposed to stay within our allocation?
This is the field in the schema which indicates the address block(s)
that lie beyond a given tunnel. This likely lies within our
allocation.
We have 3ffe:1ce1:0::/48
A tunnel could get a single subnet..
3ffe1:1ce1:0:b5::/64
or a bunch:
i.e., 3ffe:1ce1:0:ff00::/62 or 3ffe:1ce1:0:ff00::/62
or it could have several blocks (e.g., because either it's in the
middle of being renumbered, or because it got allocated several
subnets which aren't aggregateable.
- Bill
