-----Original Message----- From: Thomas Narten [mailto:narten@us.ibm.com] Sent: Wednesday, April 12, 2006 5:20 AM To: Davis, Terry L Cc: Bound, Jim; Tony Hain; PPML; address-policy-wg@ripe.net; Richard Jimmerson; Latif Ladid ("The New Internet based on IPv6"); ollivier.robert@eurocontrol.fr; Brig, Michael P CIV DISA GES-E; Pouffary, Yanick; Green, David B RDECOM CERDEC STCD SRI Subject: Re: Question - Aviation
"Davis, Terry L" <terry.l.davis@boeing.com> writes:
PSS: Back to "critical infrastructure" networks a moment, I'd say
Thomas It might; it does seem to meet requirements. We would need to spend some more time thinking about any potential impacts of address collision and how this would work DNS, gateway nodes, etc. Take care Terry that
any network that wanted to declare itself "critical infrastructure" could obtain PI space,
Note: in my mind "PI" space is closely associated with the notion of "being routed within the DFZ of the public internet".
BUT to me this type of network should always be run as a "closed network" with exchanges to the Internet only through "mediation gateways" operating at the application level, not at the routing level.
So, this type of network isn't connected directly to the internet and is thus not really part of the public internet (which makes sense to me). Thus, it is unclear to me that PI space is really needed for this.
Seems to me, that all you really need is globally unique, unrouted (on the public internet) space.
Would RFC 4193 "unique local addresses" satisify the need?
And if your answer is "they are not unique enough", would centrally assigned ones, ala (expired) draft-ietf-ipv6-ula-central-00.txt meet your needs? (It would be ironic if you answered yes, because the topic of resurrecting this document came up during the discussion of http://www.arin.net/policy/proposals/2006_2.html at Tuesday's ARIN meeting).
Thomas