Re: [dhcwg] 3315bis (RFC8415)

It is a good point, but this concerns that portion of the delegated prefix that was
not named in a prefix exclude option. In the simplest case, for a /64 delegation
there is no prefix exclude and it is up to the requesting router to manage the
prefix appropriately without breaking anything. That is where the informative
text offered in 'draft-templin-v6ops-pdhost' provides guidance.

I think it would be useful for the draft to cite and discuss RFC6603, however,
so your point is well taken.

Thanks - Fred 

> -----Original Message-----
> From: dhcwg [mailto:dhcwg-bounces@ietf.org] On Behalf Of Roy Marples
> Sent: Tuesday, December 18, 2018 2:38 PM
> To: dhcwg@ietf.org
> Subject: Re: [dhcwg] 3315bis (RFC8415) - Thanks to all!!
> 
> Isn't that what Prefix Exclude is for?
> 
> https://tools.ietf.org/html/rfc6603
> 
> Roy
> 
> On 18/12/2018 19:48, Templin (US), Fred L wrote:
> > Hi Bernie,
> >
> > We have such an operational document aligned with v6ops:
> >
> > https://datatracker.ietf.org/doc/draft-templin-v6ops-pdhost/
> >
> > Here is the text that speaks to this subject (where “the node” refers
> >
> > to the DHCPv6 PD requesting router/client and “the prefix” refers to
> >
> > the delegated prefix):
> >
> >     “In a second alternative, the node can assign as many addresses as it
> >
> >     wants from the prefix to the upstream interface over which the prefix
> >
> >     was received, but in normal practice does not assign the prefix
> >
> >     itself (or subnets from the prefix) to the upstream interface.  If
> >
> >     the node assigned the prefix to the upstream interface, any neighbors
> >
> >     on the upstream link receiving an RA could configure addresses from
> >
> >     the prefix and a default route with the node as the next hop.  This
> >
> >     could create a loop where upstream link neighbors send packets to the
> >
> >     node which in turn forwards them to another upstream link neighbor.
> >
> >     Still, there may be cases where the node provides services for
> >
> >     dependent neighbors on the upstream link that have no other means of
> >
> >     connecting to the network.  ([RFC8415] chose to remain silent on this
> >
> >     subject since it is operational rather than functional in nature.)”
> >
> > How does this look?
> >
> > Thanks - Fred
> >
> > *From:* Bernie Volz (volz) [mailto:volz@cisco.com]
> > *Sent:* Monday, December 17, 2018 5:00 PM
> > *To:* Templin (US), Fred L <Fred.L.Templin@boeing.com>
> > *Cc:* dhcwg@ietf.org; Ralph Droms <rdroms.ietf@gmail.com>
> > *Subject:* Re: 3315bis (RFC8415) - Thanks to all!!
> >
> > Thanks ... it was a major effort by many - including you!
> >
> > Regarding your questions, I believe it is the “Or, is it the case that
> > RFC8415 simply chose to remain silent on the subject and leave it to
> > other documents of a more operational nature to deal with the subject?”
> >
> > - Bernie
> >
> >
> > On Dec 17, 2018, at 7:56 PM, Templin (US), Fred L
> > <Fred.L.Templin@boeing.com <mailto:Fred.L.Templin@boeing.com>> wrote:
> >
> >     Bernie,
> >
> >     Congratulations on seeing this important work through to successful
> >     publication!
> >
> >     I do have one question regarding prefix delegation: in RFC3633, it says:
> >
> >         “Upon the receipt of a valid Reply message, for each IA_PD the
> >
> >         requesting router assigns a subnet from each of the delegated
> >
> >         prefixes to each of the links to which the associated interfaces are
> >
> >         attached, with the following exception: the requesting router MUST
> >
> >         NOT assign any delegated prefixes or subnets from the delegated
> >
> >         prefix(es) to the link through which it received the DHCP message
> >
> >         from the delegating router.”
> >
> >     Unless I missed it, I do not see the same “MUST NOT” in RFC8415.
> >     What can
> >
> >     we infer from this? Does it mean that it is now OK for the RR to
> >     assign the
> >
> >     prefix to the upstream link? Or, is it the case that RFC8415 simply
> >     chose to
> >
> >     remain silent on the subject and leave it to other documents of a more
> >
> >     operational nature to deal with the subject?
> >
> >     Thanks in advance,
> >
> >     Fred
> >
> >     *From:* dhcwg [mailto:dhcwg-bounces@ietf.org] *On Behalf Of *Bernie
> >     Volz (volz)
> >     *Sent:* Wednesday, November 28, 2018 12:02 PM
> >     *To:* dhcwg@ietf.org <mailto:dhcwg@ietf.org>
> >     *Cc:* Ralph Droms <rdroms.ietf@gmail.com <mailto:rdroms.ietf@gmail.com>>
> >     *Subject:* [dhcwg] 3315bis (RFC8415) - Thanks to all!!
> >
> >     Hi:
> >
> >     It was great to finally see the multi-year effort on 3315bis be
> >     published (RFC8415). draft-dhcwg-dhc-rfc3315bis-00 was published
> >     January 2014. My guess is that this has taken about 5 years as work
> >     began a few months prior to convert RFC3315 text into an XML to
> >     produce this first document.
> >
> >     Thanks to the co-authors, Ralph Droms (as document shepherd), the
> >     many reviewers for their feedback, and the WG in supporting the
> >     effort. And, to the RFC-Editor in their work to finalize the document.
> >
> >     But … We’re not done yet, as hopefully we’ll work to advance DHCPv6
> >     to full standard – hopefully that will be a much smaller (and
> >     shorter) effort!!
> >
> >     If you do find issues or areas of concern as you reference the
> >     document (and please start using it instead of 3315!), please be
> >     sure to file an errata or at least drop the co-authors a note.
> >
> >     Again, thanks for everyone’s efforts!!
> >
> >     -Bernie Volz
> >
> >
> > _______________________________________________
> > dhcwg mailing list
> > dhcwg@ietf.org
> > https://www.ietf.org/mailman/listinfo/dhcwg
> >
> 
> _______________________________________________
> dhcwg mailing list
> dhcwg@ietf.org
> https://www.ietf.org/mailman/listinfo/dhcwg