Re: [spring] Network Programming - Penultimate Segment Popping

[Robert Raszuk <[e-mail]>]

Tom,

> If intermediate hosts in the routing list are able to add or remove SRH
per RFC8200

"Routing list" contained in SRH does not matter at all here in terms of
compliance with RFC8200.

At each segment midpoint the outer IPv6 destination is *rewritten*. For
illustrative purposes you may visualize like outer IPv6 header is removed
and new IPv6 header is added with new destination address at each segment
midpoint.

It is this principal of SRv6 operation which is being used and which in no
way contradicts nor stretches RFC8200 as it is written today.

Besides it is basic thing that SR segment may end in the middle of the
network then native routing will take the packet to the dst. Is that also
being considered violation of something ?

Your analogy to NAT is pretty false as NAT box is never an intended
destination of the packet.

All the best,
R.


On Sat, Dec 7, 2019 at 5:27 PM Tom Herbert <[e-mail]> wrote:

> On Sat, Dec 7, 2019 at 7:10 AM Darren Dukes (ddukes) <[e-mail]>
> wrote:
> >
> > Ron, you say
> > >>  RFC 8200 addresses extension header insertion and deletion
> identically, in the same sentence.
> >
> > This sentence you refer to clearly permits PSP as defined in network
> programming:
> >    Extension headers (except for the Hop-by-Hop Options header) are not
> >    processed, inserted, or deleted by any node along a packet's delivery
> >    path, until the packet reaches the node (or each of the set of nodes,
> >    in the case of multicast) identified in the Destination Address field
> >    of the IPv6 header.
> >
> > Clearly we process the SRH at the node identified in the destination
> address field of the IPv6 header.
> > With PSP we remove the SRH at the node identified in the destination
> address field of the IPv6 header.
> >
> Darren,
>
> I believe you are assuming a very liberal interpretation of the
> wording in RFC8200. If intermediate hosts in the routing list are able
> to add or remove SRH per RFC8200, then they are allowed to add or
> remove any and all extension headers per the same interpretation. I do
> not believe that is at all the intent of RFC8200. Also note that the
> liberal interpretation it would validate other cases of arbitrary
> extension header manipulation like in NAT devices for instance.
>
> Tom
>
> > I think it’s clear we can conclude this thread.
> >
> > Darren
> >
> >
> >
> > On Dec 6, 2019, at 5:01 PM, Ron Bonica <[e-mail]> wrote:
> >
> > >> I have observed, in your original post, the conflation of SRH
> insertion within an SR Domain with the PSP behavior defined in network
> programming.
> > >> Whether this was intentional or not, I do not know.
> > >> Regardless, it is wrong.
> >
> > Darren,
> >
> > We clearly disagree.  RFC 8200 addresses extension header insertion and
> deletion identically, in the same sentence.
> >
> >
> >                                                        Ron
> >
> >
> >
> >
> >
> >
> > Juniper Business Use Only
> >
> >
> > --------------------------------------------------------------------
> > IETF IPv6 working group mailing list
> > [e-mail]
> > Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> > --------------------------------------------------------------------
>
> _______________________________________________
> spring mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/spring
>