Minutes IETF113: intarea
minutes-113-intarea-04

IntArea Ð IETF 113, Vienna, 3/22/2022

Chairs:
Juan Carlos Zuniga (Cisco)
Wassim Haddad (Ericsson)
All presentations are individual, no working group document.

No changes to the agenda

Agenda Bashing, WG & Document Status Updates (Chairs)
5 minutes

IPv4 policing at the IETF - Seth Schoen
draft-schoen-intarea-ietf-maintaining-ipv4 - 15 min

Seth: present the first two of the series of 4 drafts on IPv4 address space
Seth: IPv4 requires maintenance, and IETF is positioned for this
Local fix is local and does not raise allocation policy questions
disagreement about Ipv4 fixes
in 2016 it was proposed that IPv4 should be declared historical, hope to see
the opposite consensus. IPv4 is still widely used and some RFC are still focus
on IPv4. Eric: can the IETF commit to something forever?, clarify IPv4 fix,
protocol, addressing? Seth: I donÕt this there has been a transition strategy,
I don't suggest that the IETF should commit to something eternal. Seth: Fix
proposed by implementors, existing RFC about allocation, addressing, our
proposal is agnostic on that kind of question, we are really focused on
existing implementations. Eric: donÕt use unused addresses just to gain one
year. Tom: How many people are actually going to help?, do we care?, this
effort goes against the development of IPv6 Seth: It seem to be a possible
point of argument, our fixes alleviate the scarcity problem Jen: Fixing bug,
but not developing anything new, I have no objection in maintenance, IÕm a bit
confused about, getting higher lever consensus will help? Seth: In our
understanding, a lot of people will say that this is not beneficial because
IPv4 is done, thatÕs why we need consensus

IP Parcels - Fred Templin
draft-templin-intarea-parcels - 15 min

Fred: a segment is the retransmission unit in case of lost, and a parcel may
contain multiple segments. In that sense, it is a packet of packets. Add a
jumbo payload option and not 0 in payload length. IP parcels are based on IP
Jumbograms, it can be supported in IPv6 and IPv4. Can we adopt the document as
a WG Item ? JCZ: need support on the list Lars: how can this increase
performance? Fred: it reduced the numbers of interrupts and the overhead of the
parcel. Lars: We already do it on the ends, we have already these efficiencies.
Fred: There is no additional efficiency over GSO? Fred: The end systems see the
same efficiency that we see with GSO GRO TCP segment offload. Because a Parcel
may contain multiple segments, end systems will see better performance than
just using GSO/GRO when the Parcel is sent intact over links that have
sufficient MTU. Also, GSO/GRO are not and cannot be standardized, whereas IP
Parcels can be. Luigi: How will the network have the ability to know how to use
the parcel. Fred: We start to send regular IP packets and then you send the
probe. If the probe succeeds you can then start using parcels. Luigi: What
happens if links in the middle support IP Parcels but the end doesnÕt. Fred:
Then, it will be like regular IP packets.

Internet addressing gaps - Luigi Iaonne 15 min - update on 2 drafts
draft-jia-intarea-internet-addressing-gap-analysis
draft-jia-intarea-scenarios-problems-addressing

Luigi: side meeting, three questions arise: (i) what features do we want, (ii)
hows the feature innovation happening, (iii) what is an address ? goal was to
bring the community to discuss addressing. Documents are a community effort, WG
adoption ? JCZ: No comments from the room, discussion continues to mailing list.

Semantic IP addressing for satellite - Lin Han
draft-lhan-problems-requirements-satellite-net-02
10 min

Lin: L3 solutions for LEO constellation. Satellite are very organized. Semantic
address. A satellite can be addressed by 3 values. IGP can be used but itÕll
bring some problems. Use hybrid solutions Gorry: Which operators or
manufacturers are calling to do this constellation interoperability work? Lin:
Answer in the problem statement. Today, there are mostly different service
providers and resources like orbits and spectrum will be limited. In the future
they may need to interoperate. Gorry: Did Starlink show some interest or are
interested in interoperability? Lin: They are testing ISL this year, but for
the moment they are only using proprietary solutions.

IP regional Internet blocking considerations, Leonard (Lenny) Giuliano
draft-giuliano-blocking-considerations

Lenny: individual work with personal views.
Eric: Thanks for specifying this.
Lenny: Discussion on blocking the internet for some specific regions to
describe impact. Intended audience is Policy makers and general public. Ted:
The Internet Society (ISOC) has already addressed this with some documents in
the past: Sudan 2019 for example. Three (3) reports have been published on this
topic, addressing policy makers:

https://www.internetsociety.org/resources/doc/2022/internet-impact-brief-how-refusing-russian-networks-will-impact-the-internet/

https://www.internetsociety.org/policybriefs/internet-shutdowns/

https://www.internetsociety.org/news/statements/2019/turn-the-internet-back-on-in-sudan-and-keep-it-on/

I suggest that the authors consider reviewing this work from ISOC and
contribute there. Important to state that this document in the current shape
should not be addressed to policy makers, as they need to have the full picture
with the political and society impact, and not only the technical side. I
discourage IntArea to adopt this, and I encourage the authors to contribute
with the valuable expertise in other fora.

Lenny: Thanks for feedback. I would like to hear from the WG. Back to the
consensus, donÕt have an answer yet. Jen: I read this as a Dr specifying ways
to hurt your neighbour. Lenny: This is definitely not the intention. The
intention is to consider consequences before trying to hurt anyone. Alissa: IÕm
one of the authors of the RFC 7724, we got similar comments because this was
also triggered by events in the past. It took a long time to publish it (beyond
the timeline of the event that was actually intended for), but we got to the
publication. You canÕt really avoid the political implications of the document,
and taken out of political context it would be misinterpreted as an
endorsement. Lenny: The intention is definitely not to endorse it to policy
makers. Rudiger: Have you reviewed on any expertise in diplomacy to see the
potential side effects of this ? How can this be a constructive document, I
feel that whatever relevant or correct information you have in the document, if
you begin this way, you encourage people to behave like this, or even an
invitation to to something like this, do you really feel the public discussion
like this?, I would urge the authors to really invoke qualified diplomatic
expertise to clarify the actual effects, historical background. Lenny: We did
not consult diplomatic experts. We tried to make it clear that we are not
supporting or endorsing this. We think this is a good Informational source.
Open question BCP vs informational. Ignoring the subject or saying nothing does
not help neither. Tom: Speaking as an individual. This is trying to be a
technical document responding to a political issue. I think youÕve written very
well, I want to draw your attention about what you say and what is the message
you actually spread. Have you seen any of the Internet Sanctions project? I
think that showing the raw information is not the best way to convey the
message. IÕll not support the document and I donÕt think it should be supported
by any IETF working group. Benjamin: I appreciate the intention. Timeline is
moving fast and it is important to take a longer time horizon. Taking this in a
technical way does not respond to the actual political question, itÕs not
possible to be neutral in this matter. More productive would be to go to an
organization that does not have to be neutral and can make a political
statement. Lenny: WeÕre not trying to advocate against or for, itÕs meant to be
technical neutral. Mallory: Tomorrow weÕll have the hrpc meeting I look forward
to talk with you about this matter, I think the effort to try to describe the
problems is an important thing. I welcome the discussion, even though I also
donÕt agree with the adoption of the document in its current form. The
censorship draft referenced in the slides is a good example, and probably could
be used. So far there have been no sanctions in similar situations in the past.
IETF is in the business of keeping things connected, which is contrary to
sanctioning.

https://datatracker.ietf.org/doc/html/draft-irtf-pearg-censorship

Lenny: For transparency, we are presenting this in hrpc not for doing WG
shopping, but to get feedback from a broader audience. Regarding WG adoption,
IntArea was the only considered WG for adoption. JCZ: Thanks for the
presentation and the discussion. Lenny, I encourage you to look at the
references and the feedback you got about the best fora to provide this type of
information. Regarding adoption, now it is clear that there is no consensus and
that the IntArea group does not believe this is something we should consider
for adoption.

On Higher Levels of Address Aggregation - Tony Li
draft-li-int-aggregation-00
15 min

CIDR log time ago, not enough aggregation
new concept: abstraction naming boundary, not at this ISP boundary. ISP
generate more specific prefix, and the abstraction boundary does the
aggregation. can this apply to organization, continental or regional? regional
is possible for IPv6. Too close to the region, no traffic engineering, too far
not interest. Dino: We thought about the proxy aggregation, inbound vs
outbound, etc. Proxy aggregation may be better at the egress. Tony: In the
interest of time, letÕs discuss this offline.

Service Routing in Multi-access Edge Computing - Zongpeng Du
draft-du-intarea-service-routing-in-mec-00

Benjamin: Private information by hashing URL should not be revealed? specially
in DNS info. Tom : How can you guarantee that the DNS does not clash into
another IP no conflict? JCZ: No more time for questions or calls. We will ask
take the call for support / adoption for the drafts on the list.