IntArea WG Agenda

IETF 121 - Hybrid meeting, Dublin (Ireland) + Online

13:00-14:30 Wednesday, November 6, 2024, Afternoon Session II (Ireland,
Dublin (GMT+1))

Chairs:
Juan Carlos Zuniga
Wassim Haddad

  1. Agenda Bashing, WG & Document Status Updates (Chairs)
    5 mins

Agenda bashed.
Status update (see slides) on WG items and expired drafts for many years
that will be declared "WG dead".

Carlos J. Bernardos (CJB): ask about one particular draft that might be
of his interest to follow up
Eric Vyncke (EV), AD: as soon as a document is adopted, the document
belongs to the WG. Please feel free to continue working on that, but
suggests to contact the original authors and cited them as contributors
if they agree.

  1. Communicating Proxy Configurations in Provisioning Domains - Tommy
    Pauly
    draft-ietf-intarea-proxy-config-02
    15 mins

Tommy Pauly (TP) presents the slides.

Josh Cohen (JC): based on the scheme, what about FTP?
TP: it does not need to know that is FTP/which scheme is about.

Yaroslav Rosomacho (YR): comments about a use case he proposed.

Dragana Damjanvic (DD): on the authentication point (slide 6) shouldn't
be in this doc. So options 1 or 2.

Mirja Kuhlewind (MK): agree, this does not belong to this document

EV: as individual contributor, no on ECH key in the PvD to avoid
duplicate information with the DNS (slide 7).

DD: agrees with EV.

JC: on very long lists (slide 8) parsing a json file is as easy as
parsing a javascript file.

  1. Using Dummy IPv4 Address and Node Identification Extensions for
    IP/ICMP translators (XLATs) - Jen Linkova

    PROBE: A Utility for Probing Interfaces
    draft-fenner-intarea-probe-clarification-02

Jen Linkova (JJ) presents the slides.

Extending ICMP for Node Identification
draft-ietf-intarea-extended-icmp-nodeid-00

JJ presents the slides.

Ron Bonica (RB): one comment, issue with the extension and the extension
header due to that the extension header does not have a length
attribute. RB has a draft to fix that.

draft-equinox-v6ops-icmpext-xlat-v6only-source-00

JJ presents the slides. This has been presented in v6ops.

20 mins

  1. Stateless Reverse Traceroute - Rolf Winter
    draft-heiwin-intarea-reverse-traceroute-stateless-03
    15 mins

Rolf Winter (RW) presents the slides.

RB: shoud we never add a new ICMP messages, because legacy middle boxes
will never let them go through.
Question about backwards compatibility.

Matt Mathis (MM): how important is to trade off an optimal design to
enable traceroute through NATs when people using this would care more
about the public Internet.

EV (no hat on): second MM and Ron. We should use extended echo. Have you
tried with v6?
RW: not yet.

MM: have you considered approaches where the state is carried on the
packet?
RB: not yet.

  1. Web Proxy Auto Discovery Next Generation - Josh Cohen
    draft-joshco-wpadng-02
    10 mins

JC presents the slides.

No questions.

  1. The Multicast Application Ports - Nate Karstens
    draft-karstens-pim-multicast-application-ports-01
    10 mins

Nate Karstens (NK) presents the slides.

Dave Thaler (DT): this is a great idea. Agree assigning 2 adjacent ports
is good. But 49150 cannot be used, was assigned back in 2017. The next
two lower ports (49148 and 49149) are available though.

Lorenzo Colitti (LC): I assumed you have checked this works. We checked
in the past and in packets were being send randomly to any of the
applications.

Max Franke (MF): is this for ASM only or also SSM?
NK: should work for SSM as well.

EV (no hat): is this the right wg, as Intarea does not do port
assignment? (could be done though).
NK: it's a bit more than doing port assignment.

Stig Venass (SV): this is a great idea.

Stuart Cheshire (SC): argument in favour of this work and how multicast
should have been done on the first place ;)

MK: for using a port number in the user part you don't need an RFC.

  1. MKA over IP - Hooman Bidgoli
    draft-hb-intarea-eap-mka-00
    10 mins

Hooman Bidgoli (HB) presents the slides.

David Lamparter (DL): this needs to progress simultaneously with the
progress at the IEEE. This should be done with IPv6 as well.

Antoine Fressancourt (FC): one key for all paths or shared for all nodes
on the path.
HB: one key per flow.

EV: I find your idea fascinating. (I didn't get the question, check the
transcript, sorry)

  1. Tentative title "Analog Blockers to wide employment of Jumbo MTUs in
    the production Internet" - Matt Mathis
    5 mins

If time permits:

  1. ICMP Extension Header Length Field - Ron Bonica
    draft-bonica-intarea-icmp-exten-hdr-len-00
    5 mins

  2. DHCPv4 Option for IPv4 routes with IPv6 nexthops - David 'equinox'
    Lamparter
    draft-equinox-intarea-dhcpv4-route4via6-00
    5 mins