Re: I-D ACTION:draft-ietf-v6ops-ent-scenarios-01.txt

[Ronald van der Pol <Ronald.vanderPol@rvdp.org>]

On Wed, Feb 11, 2004 at 23:45:05 +0100, Ronald van der Pol wrote:

> I will read it tomorrow and comment
> on it. Sorry for the confusion.

General:

I find it hard to comment in detail. Basically, the document is an
extensive (and good) list of all the issues that must be handled
when introducing IPv6 in an enterprise. However, many items are not
"the business of the IETF". There are also few requirements written
down. Much depends on how the analysis document handles all the
issues. I think the focus should go now to the analysis.

section 3.2:

s/IPv6 address ownership/IPv6 address assignment/

In the 'Network Infrastructure Component 3" is says:
 - Other services (e.g. NTP, etc.........)
Those ..... look like something to be filled in. Please remove in
final version.

section 3.3:
_maybe_
s/Example Network A:/Example Network A (base scenario 1):/
etc.

sections 4.1 to 4.3 seem to be analysis instead of requirements.
I would like to see a preference for dual-stack and v4-in-v6
tunnels instead of translation, but this is something for the
analysis document. Maybe sections 4.1 to 4.3 could mention
the cases: node is v4-only, dual stack or v6-only and network
is v4-only, dual stack or v6-only?

section 5.8
Has also quite some analysis already. An interesting point would
be: is there a requirement for v4-v6 multicast interoperability?
If so, that could be a recommendation to the IETF.

I miss thing like: renumbering, stable addresses and things like
that. I thought ipv6 is working on non-routable unique addresses
because of the strong demand from enterprises. Maybe a section
about this could be added. (yes, I know; easier said than done :-)

	rvdp