[10:14:00] --- LOGGING STARTED
[12:59:30] --- LOGGING STARTED
[13:03:30] --- LOGGING STARTED
[16:14:03] --- j.schoenwaelder@jabber.eecs.iu-bremen.de has joined
[16:15:12] --- bkhabs@jabber.org has joined
[16:17:15] --- jc_lee has joined
[16:19:43] --- Carlos P. has joined
[16:24:46] --- nm has joined
[16:26:16] --- brabson has joined
[16:27:18] --- ogud has joined
[16:27:32] --- nm has left: Replaced by new connection
[16:27:33] --- nm has joined
[16:27:33] --- nm has left
[16:27:50] --- sivaram has joined
[16:28:24] --- brabson has left
[16:29:18] --- nm has joined
[16:29:53] --- cabo--tzi--org@jabber.org has joined
[16:30:51] --- brabson has joined
[16:33:17] --- brabson has left
[16:33:30] --- brabson has joined
[16:33:35] --- sleinen has joined
[16:38:19] --- petrescu7 has joined
[16:38:51] --- momose has joined
[16:38:54] <petrescu7> anybody in this jabber room who is not in the intarea meeting room at Hilton?
[16:39:05] <bkhabs@jabber.org> Yes.
[16:39:08] <Carlos P.> Yes
[16:39:12] <j.schoenwaelder@jabber.eecs.iu-bremen.de> yes
[16:39:46] --- loughney has joined
[16:39:56] <loughney> WG news -
[16:39:59] <petrescu7> PEople who are not at Hilton get mp3 streaming?
[16:40:10] <loughney> SHIM6 - taling about intra-site TE
[16:40:10] --- dthaler has joined
[16:40:19] <loughney> L*VPN - security and multicast
[16:40:22] <bkhabs@jabber.org> Yes, I am receiving the mp3 feed.
[16:40:30] <loughney> PWE3 -multisegment pseudowires
[16:40:43] <loughney> & mpls/IP psuedowires
[16:40:48] <loughney> SOFTWIRE WG - had interim meeting
[16:40:55] <loughney> Drafts being discussed
[16:41:03] <loughney> - draft-fenner-iana-exp-2780
[16:41:05] <petrescu7> Jari presents two drafts which have no WG, but are in INT area
[16:41:17] <loughney> - draft-bagnulo-cga-ext-01.txt
[16:41:49] <loughney> (do I still have connectivity)
[16:42:15] --- Yoshifumi Atarashi has joined
[16:42:20] <loughney> - draft-bonica-internet-icmp-01.txt
[16:42:20] <petrescu7> other: draft-bonica-internet-icmp and draft-bagnulo-ipv6-rfc3480-update
[16:42:34] <loughney> - draft-bagnulo-ipv6-rfc3480-update-00.txt
[16:42:57] --- dthaler has left: Replaced by new connection
[16:43:29] --- nm has left: Replaced by new connection
[16:43:29] --- nm has joined
[16:43:29] --- nm has left
[16:43:44] <petrescu7> draft-laganier-ipv6-khi
[16:43:58] <loughney> - draft-laganier-ipv6-khi-01 - needed for HIP address to avoid referral problems. had had past discussions.
[16:44:19] --- dthaler has joined
[16:44:19] <petrescu7> if you have any comments, or if I forgot any documents please feel free mike
[16:44:51] <loughney> Alex, do you want to jabber? my network connection is a bit flakey
[16:44:55] <petrescu7> draft-fenner-iana-exp-2780
[16:45:00] <petrescu7> John I can jabber.
[16:45:40] <petrescu7> MArgaret speaking about experimental allocations for...
[16:47:12] --- frodek has joined
[16:47:32] <petrescu7> slide: Status
[16:48:33] <petrescu7> MW asking for questions/comments
[16:48:44] <petrescu7> anybody not at Hilton can ask, I could relay to mike.
[16:48:53] <petrescu7> slide: Common IPv6 Issues
[16:49:19] <petrescu7> slide: Motivation
[16:50:26] <petrescu7> slide: common IPv6 "Gotchas"
[16:51:05] <petrescu7> slide: Address Architecture
[16:52:28] <petrescu7> Bob Hinden questioning about some wrong address.
[16:52:54] --- dthaler has left
[16:53:00] <petrescu7> slide: Scoped Unicast Addressing
[16:54:29] <petrescu7> slide: MTU and Fragmentation
[16:55:05] <petrescu7> IPv6 packets do not fragment packets
[16:56:20] <petrescu7> slide: ND vs. ARP
[16:57:37] <petrescu7> Dave Thaler: unlike IPv4 ARP is only on Ethernet, ND is performed on anything
[16:57:55] <petrescu7> slide: IPv6 Documentation Addresses
[16:58:13] <petrescu7> slide: IPv6 Required?
[16:58:55] <petrescu7> Pekka Nikander approaching mike.
[16:59:12] <petrescu7> sorry that's not Pekka Nikander but... (?)
[16:59:38] <petrescu7> Pekka Savola: I'm not sure this is policy, but there's a guideline doucment.
[16:59:39] --- Tsubasa has joined
[16:59:55] <petrescu7> speaker: should discuss in the Charter, not as last minute.
[17:00:19] <petrescu7> speaker: at some point before go to IESG it should be clarified.
[17:00:26] --- loughney has left
[17:00:58] <petrescu7> speaker: shim6 specifically charter says no IPv4
[17:01:11] <petrescu7> speaker: as part of IESG Charter review should be made apparent
[17:01:22] --- loughney has joined
[17:01:49] <petrescu7> speaker: jean-francois mlle? in our charter we specifically mention IPv6, but by the time there were systems not capable of IPv6.
[17:01:54] --- jinmei has joined
[17:02:12] <petrescu7> speaker: even if it's in scope there may be parts of MIB(?) actually relaxing
[17:02:43] <petrescu7> speaker: Tim Chown, Charter, IESG should be there.
[17:02:51] <petrescu7> TC: IPv6 considerations section?
[17:03:14] <petrescu7> Bob Hinden: Charter is the right place, there you know what the WG is doing.
[17:03:57] --- ahamza has joined
[17:04:08] <petrescu7> John Loughney: question: multicast?, I guess there wshould be a general assumption, if the Charter doesn't say explicitely v4/v6 then what is the correct assumption?
[17:04:46] <petrescu7> Pekka Savola: there was a document IETF Specifications lack of IPv6 support, we should not do redo that document.
[17:05:01] <petrescu7> PS: should be required by default.
[17:05:43] <petrescu7> Jari: problem of cable?
[17:05:44] --- rvdp has joined
[17:05:50] <petrescu7> Ralph Droms is going to present.
[17:06:18] <petrescu7> anyone who thinks I'm too verbose please let me know, thanks.
[17:06:35] <petrescu7> slide: IPv6 in CableLabs DOCSIS 3.0
[17:07:22] <petrescu7> slide: Agenda
[17:08:01] <petrescu7> details in http://www.comcast.com/downloads/ietf/draft-mule-cablelabs-docsis3-ipv6-00.txt
[17:08:03] --- ka has joined
[17:09:09] <petrescu7> speaker: url is actually cablelabs.com not comcast.com
[17:09:39] <petrescu7> Dslide: Motivation: why IPv6 in cable networks
[17:10:44] <petrescu7> Ralph: "MSO" is a "cable operator".
[17:12:40] <petrescu7> slide: DOCSIS 3.0 Modems and Mode of Operation
[17:14:30] <petrescu7> slide: CM Features
[17:14:47] <petrescu7> slide: back to DOCSIS 3.0...
[17:16:02] <petrescu7> slide: CM Features
[17:17:52] --- bkhabs@jabber.org has left: Logged out
[17:18:41] <petrescu7> slide: picture of topology of access model, CMTS router, Servers, says: DOCSIS 3.x IPv6 Example Architecture
[17:20:04] <petrescu7> Dave Thaler: how is the CM typically getting an address? (ahead of time or protocol?)
[17:20:21] <petrescu7> Ralph: DHCPv6
[17:20:43] <petrescu7> slide: address configuration choices
[17:20:50] --- yjs has joined
[17:22:10] <petrescu7> Dave Thaler: in a first of the three models (bridge) filtering can't be done there right?
[17:22:22] <petrescu7> Ralph: the CPEs also have to use DHCPv6
[17:22:41] <petrescu7> Jim Bound HP: privacy?
[17:22:52] <petrescu7> JB: market entities mandate privacy
[17:22:53] --- yjs has left: Lost connection
[17:23:12] <petrescu7> JB: rfc3041, overhead for stateless, complex to DHCPv6,
[17:23:31] <petrescu7> JB: if I change the EUI and I got a DHCPV6 server how do we make sure DHCPv6 data is the same?
[17:23:57] <petrescu7> Ralph: quickly saying, it's not possible for the client to change its address if perform DHCPv6
[17:24:06] <petrescu7> Ralph: I'll think about anonimising.
[17:24:19] <petrescu7> Ralph: anonymity is needed if move to places, but here we got that already.
[17:24:31] <petrescu7> Ralph: the address aassigned by dhcpv6 is not eui-64 based.
[17:24:47] <petrescu7> JB; devices are not moving yes, but I may not want my ISP have knowlkedge of the lower EUI-64
[17:24:52] <petrescu7> Ralph: then find another ISP.
[17:25:05] <petrescu7> Margaret: cut the line
[17:25:28] <petrescu7> Dave: one purpose of 3041 is if osmeodby moves, other purpose is cracking over time somebody staying in the same place.
[17:25:43] <petrescu7> Dave: for that purpose use temopporary address, DHCPv6 does it.
[17:26:06] <petrescu7> Pekka Savola: 300 million lines of something?
[17:26:15] <petrescu7> Ralph: track addresses and write them down and track?
[17:26:27] <petrescu7> slide: Conclusion
[17:27:53] <petrescu7> Margaret: Dave Thaler to present
[17:28:19] <petrescu7> agenda is at http://www3.ietf.org/proceedings/06mar/agenda/intarea.txt
[17:28:33] <petrescu7> slide: issues with protocols proposing multilink subnets
[17:28:41] <petrescu7> draft-thaler-something
[17:28:46] <petrescu7> slide: definitions
[17:29:25] <petrescu7> slide: current IP model
[17:29:37] <petrescu7> rfc 1884 2373 3513
[17:30:29] <petrescu7> slide: Internet Area seems to be fragmenting
[17:32:01] <petrescu7> slide: why did the IPv6 WG reject Multilink subnets
[17:32:54] <petrescu7> slide: Duplicate Address Detection
[17:32:59] --- miyahiro has joined
[17:33:44] <petrescu7> slide: TTL/Hop Limit issues
[17:35:08] --- miyahiro has left: Replaced by new connection
[17:37:16] <petrescu7> slide: security issues
[17:38:46] <petrescu7> slide: link-scoped multicast/broadcast
[17:40:37] <petrescu7> Thomas NArten approaching mike
[17:40:57] <petrescu7> TN: last bullet, lack of multicast break the IP model?
[17:41:15] <petrescu7> TN: either works or doesn, different than saying "works partially sometimes".
[17:41:38] <petrescu7> Dave: in IETF we tend to rely on link multicast working, otherwise...
[17:41:54] <petrescu7> Dave: NBMA is generally a problem.
[17:42:01] <petrescu7> slide: Discusssion
[17:42:26] <petrescu7> sllide shoulwd we A sticj to the IP model, B change the IP model (we as an Area)
[17:42:37] <petrescu7> C continue fragmenting
[17:42:57] --- nm has joined
[17:43:11] <petrescu7> TN: approach, job security or dead end?
[17:43:16] <petrescu7> TN: I advocate A
[17:43:31] --- momose has left: Replaced by new connection
[17:43:34] <petrescu7> TN: use MIP6 as example, the degre to which they go down , fall down of the cliff, how much does it matter?
[17:43:40] <petrescu7> TN: problem in practice?
[17:43:44] <petrescu7> TN: don't go there.
[17:43:54] --- momose has joined
[17:44:16] <petrescu7> Dave: MIIP6 specified the HA says that wehn you have a HoA remote, oforwarding packets you decrement TTL, MN sends to a multicast address (across the tunnel)
[17:44:30] <petrescu7> TN: this breaks that the HomeLAn draft.
[17:44:41] <petrescu7> Sorry that's Dave Thaler not TN.
[17:44:48] <petrescu7> Dave: explains why breaks.
[17:44:54] <petrescu7> Hesham Soliman approaches mike.
[17:45:15] <petrescu7> HS: problem is that there are 2-3 classes problems.
[17:45:28] <petrescu7> HS: explains A, B C.
[17:46:04] <petrescu7> HS: MIP specifically uses tunnel, for that the specific issues can be removed in a way, a tunnel is well documented.
[17:46:12] <petrescu7> HS: the problem is still local multicast.
[17:46:18] <petrescu7> HS: is it a problem that should happen?
[17:46:41] <petrescu7> HS: it may be a problem that can be rectified by making people use global address on a link.
[17:47:05] <petrescu7> Dave: doing update to MIP6, don't decrement ttl, allow link-scope multicast ttl, update is probably easy.
[17:47:44] <petrescu7> HS: even so, the problem is in general multicast-related and link-scope issue. What's fundamentally unique about that, why so many applications use link-lcaols.
[17:48:16] <petrescu7> Dave: some class of these are intended to be used in the context of a house. hopw fast can I get to market.
[17:48:38] --- miyahiro has joined
[17:48:38] <petrescu7> Dave: a different class, use to exchange information bootstrapt
[17:48:45] <petrescu7> MArgaret Wasserman cuts line somewhere.
[17:48:55] <petrescu7> Dave: RIP exchanges and routing protocols
[17:49:10] <petrescu7> HS: for the home doesn't work to have multiple subnets.
[17:49:21] <petrescu7> Dave: works for most of customers, desirable for many compnaniese.
[17:49:35] <petrescu7> HS: maybe there's another path, in addition to those three
[17:49:41] <petrescu7> Dave: keep thinking about
[17:50:00] <petrescu7> speaker: some situations where there's no other possibility, in MANET, multiple links.
[17:50:06] <petrescu7> Dave: I'd disagree with that.
[17:50:16] <petrescu7> speaker: multiple links, administratively...
[17:50:24] <petrescu7> speaker: possible to avoid DAD if ...
[17:50:31] <petrescu7> Dave: yes, but that's only one problem.
[17:50:47] <petrescu7> MW: awsome talk. we should not loose the benefits of the discussion.
[17:51:01] <petrescu7> MW: I would pick A, except when A is not the right choice.
[17:51:14] --- momose has left: Replaced by new connection
[17:51:25] --- rbless has joined
[17:51:30] --- ka has left
[17:51:30] <petrescu7> MW: I would love to see us doucment this, is the model to document people, realize making an exception. Not some accidentally fall of that...
[17:51:40] <petrescu7> MW: it was really a mistake, we repeat the learnings of the past.
[17:51:54] <petrescu7> Dave: multiple WGs do this independently...
[17:52:30] <petrescu7> Jari Arkko: task, I understand what is the scope of this problem, you have an example. Where are the other cases where this problem appears. VPN Gateways.
[17:52:48] <petrescu7> Dave: discussed in DNA, impacted in netlmm, autoconf for manet, mip6. Independent examples.
[17:53:13] <petrescu7> Greg Daley: we whosuled be careful regarding the multicast.
[17:53:44] <petrescu7> GD: constrained bandwiodth
[17:54:05] <petrescu7> Dave: one common solution is the concept of an IGMP snooping switch.
[17:54:15] <petrescu7> GD: we push the problem down the...
[17:54:20] <petrescu7> GD: is this the right solution?
[17:54:32] <petrescu7> Dave: yes I agree with you.
[17:54:49] --- loughney has left
[17:54:58] <petrescu7> speaker: definitely good to publish this. but past history, having another draft,...
[17:55:00] --- loughney has joined
[17:55:07] <petrescu7> speaker: make a decision about A B or C.
[17:55:12] --- TJ has joined
[17:55:29] <petrescu7> Dave: what do you Jari think the next shotop be
[17:56:00] <petrescu7> Jari: we need to document this issue in more detail and discuss in here, it doesn't really so much if you focus affecting these WGs. One more round of understanding the problem better.
[17:56:14] <petrescu7> MW: how many people read the Dave draft.
[17:56:24] <petrescu7> MW: discuss this on the int-area mailing list.
[17:56:40] <petrescu7> Mark Townsley next speaker.
[17:56:56] <petrescu7> slide: "Routing PRotocol Standarization Criteria"
[17:57:50] <petrescu7> slide: a note about running code
[17:58:00] <petrescu7> slide: rfc1264
[18:00:08] <petrescu7> slide: rfc1264 motivation
[18:00:35] <petrescu7> slide: rfc1264 - current impact
[18:00:57] <petrescu7> draft-fenner-zinin-rtg-standard-reqts-01.txt
[18:03:52] <petrescu7> slide: Process Summary
[18:04:27] --- brabson has left
[18:05:04] --- miyahiro has left: Replaced by new connection
[18:05:35] <petrescu7> talk between Mark and Bill
[18:05:50] <petrescu7> slide: obvious questions for the INT-AREA
[18:06:16] --- brabson has joined
[18:07:05] <petrescu7> slide: will this slow down my draft getting to RFC?
[18:08:12] <petrescu7> slide: will our documents improve?
[18:10:51] <petrescu7> slide: what about my document?
[18:12:42] <petrescu7> Mark TOwnsley is MT, inviting for comments.
[18:13:06] --- rbless has left
[18:13:28] <petrescu7> TN: concern, the workld has changed significantly; if an ISP wants to deploy an IGP, let them.
[18:13:51] <petrescu7> TN:other thing, broken (if not mentinons IPv6).
[18:14:04] --- brabson has left
[18:14:05] <petrescu7> TN: wondering, is that really happening today? Or some scare tactics.
[18:14:29] <petrescu7> MT: it is happening to me now. There is various procedure for that. With respect to can we make exceptions for IPv6? YEs we can? Free? No.
[18:14:48] <petrescu7> MT: because I have to or somebody has to decide. Does it fall under this criteria or not, being dair.
[18:14:53] <petrescu7> fair not dair.
[18:15:01] <petrescu7> Bob Hinden: folks in RT area hate this.
[18:15:13] --- inet6num@jabber.org has joined
[18:15:35] --- brabson has joined
[18:15:41] <petrescu7> BH: the world is really different. the reasons causing writing this. We see potentially new many routing protocols, we thought it couldn't work, afraid of serious harm.
[18:15:51] <petrescu7> BH: made it harder to create protocol standards.
[18:15:59] <petrescu7> BH: I'm not sure it is necessary as it was.
[18:16:14] <petrescu7> BH: the system now, where we don't care about implementations... maybe think about more.
[18:16:28] <petrescu7> BH: we can get crazy being very detailed oriented what's in the implementation reports.
[18:16:38] <petrescu7> BH: the BGP reports where amazing, was it necessary.
[18:16:42] --- momose has joined
[18:16:48] <petrescu7> BH: the intent was for new rt protocols.
[18:17:00] <petrescu7> BH: should be applied with some analysis, not just jack-off..
[18:17:05] <petrescu7> MW: 4 minutes left.
[18:17:17] <petrescu7> Alex Zinin: world has changed definitely.
[18:17:47] <petrescu7> AZ: primari motivation changed from preventing routing meltouts towards specs in std track contain enough info, proerply specified so that interoperable implementations.
[18:17:54] <petrescu7> AZ: this is not a theoretical problem at all.
[18:18:08] --- loughney has left
[18:18:14] <petrescu7> AZ: I saw documents that with simple questions, I feel I'mn the very person looked at it.
[18:18:30] <petrescu7> AZ: there are also very complex documents, that there were many things that were not quite right.
[18:18:41] <petrescu7> AZ: some of the features were nt' implementated at all.
[18:18:45] --- ogud has left
[18:18:57] <petrescu7> AZ: main point make sure implementations happen with enough detail, the priority.
[18:18:59] --- brabson has left
[18:19:15] <petrescu7> DT: reading table on slide.
[18:19:30] <petrescu7> DT: IPv6 question becomes a subjective thing
[18:19:35] <petrescu7> DT: in this context...
[18:20:04] <petrescu7> DT: I wd classify IPv6 not a major feature but goes into. Two implementations is a major features.
[18:20:27] <petrescu7> DT: lots of rt protocols just say use "just IPsec". It turned out you can't "just use IPsec".
[18:20:44] <petrescu7> DT: adverse effects of just saying use IPsec.
[18:20:50] <petrescu7> DT: document that kind of case.
[18:21:09] <petrescu7> speaker: wonder whether major is the right word for feature.
[18:21:25] <petrescu7> speaker: IPv6 may be fvery important but not very risky.
[18:21:29] <petrescu7> MW: why risky
[18:22:00] <cabo--tzi--org@jabber.org> (speaker == Ross Callon)
[18:22:03] <petrescu7> speaker: some value in a field (1-bit prefix lengths). You're not really worried... make any sense? 1-bit prefix is not a major feature.
[18:22:06] --- brabson has joined
[18:22:20] <petrescu7> Ross Callon: the rules can't be perfect.
[18:22:33] <petrescu7> RC: not sure if you necessarily if every case last call ...
[18:22:59] <petrescu7> MT: all those comments, don't feel like repeating yourself you say it to rt area meeting, make your oppinions there.
[18:23:18] <petrescu7> MT: risky, significiant, enforce criteria, but at the end of the day we have to use a judgement.
[18:23:41] <petrescu7> MT: to AZ, how impolememntation reports improve quality. I totally agree they do. That's why we have.
[18:24:01] <petrescu7> MT: do we allow ourselves to use a lot of judgement to freeze the proposed standard and then really clean it up or
[18:24:15] <petrescu7> MT: or we say we let it sit in ID and... That's the distinction for me.
[18:24:20] <petrescu7> Alain Durand
[18:24:26] --- rvdp has left
[18:24:33] <petrescu7> AD: judgement call should ebe a contract between IESG and the WG.
[18:24:34] --- sivaram has left
[18:24:41] --- TJ has left
[18:24:46] --- momose has left
[18:24:54] <petrescu7> MT: if we do this we should define it upfront to the WG, it really affects WG operates.
[18:24:59] <petrescu7> John Loughney at mike.
[18:25:07] --- brabson has left
[18:25:16] <petrescu7> JL: I noticed that first implementations tend to be done by students.research/
[18:25:20] --- Tsubasa has left
[18:25:29] <petrescu7> JL: even if actually what gets deployed is something else.
[18:25:41] <petrescu7> JL: actually you may not really see what's done.
[18:26:01] <petrescu7> MT: when you have stud/res produce implementations you operate in different mode. Not with BGP happened.
[18:26:17] <petrescu7> MT: all good points please repeate the m in next meeting.
[18:26:22] <petrescu7> looks like people leave the room by packs.
[18:26:29] <petrescu7> adjourned I believe.
[18:26:40] --- inet6num@jabber.org has left
[18:26:55] --- jc_lee has left
[18:27:16] --- petrescu7 has left
[18:31:15] --- jinmei has left
[18:31:15] --- nm has left: Replaced by new connection
[18:31:15] --- nm has joined
[18:31:16] --- nm has left
[18:33:57] --- cabo--tzi--org@jabber.org has left: Logged out
[18:41:20] --- frodek has left
[18:42:10] --- ahamza has left
[18:43:22] --- sleinen has left
[18:43:43] --- j.schoenwaelder@jabber.eecs.iu-bremen.de has left
[18:49:18] --- Yoshifumi Atarashi has left: Replaced by new connection
[19:13:13] --- Carlos P. has left: Logging off
[19:19:15] --- ogud has joined
[19:26:24] --- sivaram has joined
[19:29:41] --- ogud has left
[19:37:20] --- sivaram has left