Hi Geoff, as I said to Steffan, there is a reference to rfc 6832 that covers interworking, that should be used as reference.
PITR should announce the largest aggregate possible, ideally the /32. Let’s try again ;-) The prefix must be used for LISP experimentation and must not be used as normal prefix. Interworking between the EID block prefix and the non-LISP Internet is done according to [RFC6832] and [I-D.ietf-lisp-deployment]. Better? Luigi On 2 Dec. 2013, at 20:56 , Geoff Huston <[email protected]> wrote: > I'm afraid thats its still unclear to me - "natively announced" is a term I > do not understand. > > Are you saying that all more specifics _and_ the aggregate /32 itself must > not be advertised into the IPv6 global unicast network? > > Or are you saying that all more specifics of this /32 must not be advertised > into IPv6 global unicast network, but the aggregate /32 should (must?) be > advertised? > > i.e. is this /32 unreachable from the rest of the IPv6 net in its entirety, > ore are you saying that any LISP gateways into this /32 must advertise the > entire /32 and not just advertise more specifics? > > > regards, > > Geoff > > > > > > > On 3 Dec 2013, at 1:37 am, Luigi Iannone <[email protected]> wrote: > >> Hi Geoff, >> >> I agree that the sentence is unclear. >> >> The purpose was to make sure that people do not get an IP prefix for free >> and inject it in BGP, so it is close your interpretation a). >> >> What about the following text: >> >> The prefix must be used for LISP experimentation and must not be >> used as normal prefix, hence, it must be used according to [RFC6832] >> and [I-D.ietf-lisp-deployment] and must not be natively announced in the >> BGP routing infrastructure. >> >> Would this solve the issue? >> >> Luigi >> >> >> >>> The prefix must not be used as normal prefix and must not be announced in >>> the BGP >>> routing infrastructure. >> >> >> >> >> On 2 Dec. 2013, at 02:40 , Geoff Huston <[email protected]> wrote: >> >>> It may be minor, but the last sentence in section 4 is unclear. The txt I >>> see is: >>> >>> >>> ---------------- >>> >>> The prefix must not be used as normal prefix and announced in the BGP >>> routing infrastructure. >>> >>> >>> >>> >>> >>> >>> in the BGP >>> routing infrastructure. >>> >>> --------------- >>> >>> There is some error in the markup processing used to generate this draft, >>> but the more >>> substantive comment is that I can't clearly parse the sentence. >>> >>> Did the authors mean: >>> >>> >>> a) The prefix must not be used as normal prefix and must not be announced >>> in the BGP >>> routing infrastructure. >>> >>> or >>> >>> b) The prefix must not be used as normal prefix but will (must?) be >>> announced in the BGP >>> routing infrastructure. >>> >>> or >>> >>> c) something else >>> >>> >>> I would prefer to see this resolved before saying "we're done!" which I >>> suppose is a >>> "NO" response to the WGLC at this stage. >>> >>> regards, >>> >>> Geoff >>> >>> >>> >>> >>> On 29 Nov 2013, at 12:39 pm, Terry Manderson <[email protected]> >>> wrote: >>> >>>> All, >>>> >>>> As requested in Vancouver, and after the authors updated the draft (to >>>> -07) based on the Vancouver in-room discussion. >>>> >>>> This starts a 14 day last call for this document, the last call will end >>>> on Friday the 13th December 2013. >>>> >>>> You will find its text here: >>>> http://www.ietf.org/id/draft-ietf-lisp-eid-block-07.txt >>>> >>>> Please review this WG item and provide any last comments. >>>> >>>> Cheers >>>> Terry >>>> _______________________________________________ >>>> lisp mailing list >>>> [email protected] >>>> https://www.ietf.org/mailman/listinfo/lisp >>> >>> _______________________________________________ >>> lisp mailing list >>> [email protected] >>> https://www.ietf.org/mailman/listinfo/lisp >> > _______________________________________________ lisp mailing list [email protected] https://www.ietf.org/mailman/listinfo/lisp
