IETF
v6ops
v6ops@jabber.ietf.org
Monday, 26 March 2012< ^ >
Patrik Halfar has set the subject to: v6ops -- IETF80
Room Configuration

GMT+0
[05:10:56] Maoke joins the room
[05:11:31] Maoke leaves the room
[06:29:50] Joel Jaeggli joins the room
[06:36:47] <Joel Jaeggli> IETF 83 v6ops Monday 0900
[06:40:35] <Joel Jaeggli> http://www.ietf.org/proceedings/83/agenda/agenda-83-v6ops.html
[06:45:15] <Joel Jaeggli> meeting materials are here:
[06:45:17] <Joel Jaeggli> https://datatracker.ietf.org/meeting/83/materials.html
[06:45:22] Maoke joins the room
[06:45:47] Maoke leaves the room
[06:46:00] Maoke Chen joins the room
[06:50:52] tskj joins the room
[06:57:02] <Joel Jaeggli> agenda notes - 6204bis discussion moved to thursday.
[06:57:56] <Joel Jaeggli> IPv6 Guidance for Internet Content and Application Service Providers <http://tools.ietf.org/html/draft-carpenter-v6ops-icp-guidance> moving to monday
[06:58:48] <Joel Jaeggli> http://tools.ietf.org/html/draft-carpenter-v6ops-label-balance - moving to monday
[07:00:30] dudisaki joins the room
[07:00:58] <Joel Jaeggli> meeting commencing
[07:01:10] <Joel Jaeggli> agenda bashing
[07:01:27] Olle E. Johansson, Paris, France for the IETF 83 (GMT +1) joins the room
[07:01:38] york joins the room
[07:01:49] Lee Howard joins the room
[07:01:50] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Hello folks
[07:01:54] Akira Kato joins the room
[07:02:25] Andrew McGregor joins the room
[07:04:42] <Joel Jaeggli> the iesg is asking about a group interim meeting right around ripe in sept. the question to this working group is we always have far to much on our agenda would you like an interim meeting 24-28th of sept
[07:04:46] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Lee Howard on mike
[07:05:10] <Joel Jaeggli> lee howard - the volume of work will expand to available capacity
[07:05:47] <Joel Jaeggli> test how many of you will be at ripe anyway?
[07:05:55] <Joel Jaeggli> about a dozen hands
[07:06:06] <Joel Jaeggli> hum for interim,
[07:06:14] <Joel Jaeggli> low hum in favor,
[07:06:15] <york> Hum was for interim... no opposition
[07:06:35] <Joel Jaeggli> presentation 1
[07:07:10] <Joel Jaeggli> http://www.ietf.org/proceedings/83/slides/slides-83-v6ops-0.pdf
[07:09:15] iljitsch joins the room
[07:09:23] Wes George joins the room
[07:09:33] <iljitsch> no power where I'm sitting at the left side of the room :-(
[07:09:59] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Lots of free power sockets on the other side
[07:10:02] kbransom joins the room
[07:10:20] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> So I guess the left side of the room symbolise the shortage of addresses in the IPv4 space
[07:10:41] <iljitsch> no, when I'm on the v6-only network I still don't get any power
[07:10:46] Atarashi Yoshifumi joins the room
[07:10:49] <Lee Howard> we have IPv4 extension cords
[07:11:07] <york> :-)
[07:11:09] <iljitsch> they stretch a bit but snap back just when you need them the most
[07:11:14] Bjoern A. Zeeb joins the room
[07:11:17] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> he he
[07:11:26] <iljitsch> do we have nat64 this week, btw?
[07:11:30] <Bjoern A. Zeeb> no
[07:11:38] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> We have an IPv6 only wifi net
[07:11:56] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> don't know if it's got NAT64
[07:11:59] Dominik Elsbroek joins the room
[07:12:01] <Bjoern A. Zeeb> no
[07:12:50] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Anyone on XMPP that is not in the room?
[07:13:28] <Wes George> yes, watching from IDR
[07:13:53] <Joel Jaeggli> slide 8
[07:14:26] arifumi joins the room
[07:15:12] <Joel Jaeggli> analysis of dns64 logs about innapropiate AAAA replies
[07:18:07] <Joel Jaeggli> slide 10
[07:19:19] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Slide 11
[07:19:39] Janos Mohacsi joins the room
[07:19:55] <Joel Jaeggli> new problems detailed in draft appendix
[07:21:20] geir joins the room
[07:21:30] naikumar@cisco.com joins the room
[07:21:56] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Who wants to/should adjust Happy Eyeballs parameters?
[07:22:21] <Joel Jaeggli> good question
[07:22:30] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Slide 12 MTU issues
[07:23:00] <Joel Jaeggli> given not all implmetations make the same choices presumably there's some variance in what is considered optimal
[07:24:08] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> So it's application constraints then?
[07:24:25] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> For SIP I can't have 30 seconds… People give up a call after just a few secs
[07:24:38] <Joel Jaeggli> fred - issues with 464xlat mtu around 1260
[07:24:46] iljitsch leaves the room
[07:24:55] jpc joins the room
[07:25:27] <Joel Jaeggli> address port dependant mapping was detected in 4rd/ppoe
[07:25:42] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Iljitch on mike
[07:26:06] <Joel Jaeggli> v6 has no fragment in the netowkr, where do you see it?
[07:26:51] <Joel Jaeggli> which kind of traffic do you need fragmentation for?
[07:27:15] <Joel Jaeggli> answer udp particularly vpn applications
[07:27:29] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Page 13
[07:28:40] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Slide 14
[07:28:54] <Joel Jaeggli> difference between bind and unbound faailure rate.
[07:29:00] Diego R. Lopez joins the room
[07:30:23] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> "Troubleshooting is very hard"
[07:30:34] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Is this something that needs more focus?
[07:31:10] <Bjoern A. Zeeb> Or more experience?
[07:31:14] <Joel Jaeggli> addresses are too long to type
[07:31:19] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> SLide 15
[07:31:24] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Open issues
[07:31:46] <Joel Jaeggli> walt mathews - alcatel lucent
[07:32:04] <Joel Jaeggli> descriptions were not enough to understand what the problem is
[07:32:14] <Joel Jaeggli> needs more elucidation
[07:32:34] <Joel Jaeggli> simon perault ?
[07:33:19] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> that was dan york on mike
[07:33:21] <Joel Jaeggli> - nat64 with dns64 discovered many implmentation problms that we had to implment many work-arounds.
[07:34:23] <Joel Jaeggli> fred - whats the long term plan for this this draft
[07:34:37] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> more of us needs to run labs like this - and report results
[07:34:45] <Joel Jaeggli> answer - another camp in sept
[07:34:56] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> http://tools.ietf.org/html/draft-donley-behave-deterministic-cgn-01
[07:34:56] <Joel Jaeggli> donnelly
[07:35:02] <Joel Jaeggli> http://www.ietf.org/proceedings/83/slides/slides-83-v6ops-1.pdf
[07:35:02] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Donley
[07:35:50] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> 23 Mbps logging traffic from ONE cgn!
[07:36:15] iljitsch joins the room
[07:36:35] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> …someone in the audience said "Excellent. It's a feature" :-)
[07:36:35] <Joel Jaeggli> indeed
[07:36:48] <Lee Howard> what's with the psychedelic color changes on the screen?
[07:36:57] <Joel Jaeggli> it's not really that much in any event
[07:37:07] <Joel Jaeggli> slides are white
[07:37:19] <Lee Howard> they look yellow from here
[07:37:28] <Joel Jaeggli> yeah
[07:37:31] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> somone from Google asking a q
[07:37:33] <Joel Jaeggli> I saw green too
[07:37:39] <Joel Jaeggli> erik kiline
[07:37:39] <Lee Howard> Erik Kline from Google
[07:38:25] <Joel Jaeggli> public safety = law enforcement
[07:38:29] <Lee Howard> Joel, did you say the volume didn't seem like much? Because I was having trouble figuring out how I might query a database that large
[07:39:04] <iljitsch> hm, doesn't BEHAVE say you need to keep the source port the same if possible? this would break that
[07:39:09] kawashimam joins the room
[07:39:17] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Should the IETF create solutions to support law enforcement better? ;-)
[07:39:27] <Joel Jaeggli> user behavior data for big facebook apps is bigger than that
[07:39:37] <iljitsch> although I guess on a CGN you're running through ports so fast you wouldn't be able to give the client the same port it picked itself anyway
[07:39:48] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Bob Matthews Alcatel
[07:39:51] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> on mike
[07:41:09] <Lee Howard> Maybe Facebook has better DBAs than I do, or more money to spend on logging systems. Any guess what the $ per user would be?
[07:41:38] S73654-TX0FB16B913 joins the room
[07:42:12] <Joel Jaeggli> I'm told that we have the worlds largest vertical cluster
[07:42:19] <Joel Jaeggli> (by hp)
[07:42:39] <Joel Jaeggli> er veritica
[07:42:48] <Lee Howard> SO, let's imagine that an ISP using CGN doesn't want to have the world's second largest cluster
[07:43:12] <Joel Jaeggli> probably your reportin requirements are different
[07:43:27] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> There's talk about creating a v4 exit group
[07:43:28] <Joel Jaeggli> next steps
[07:43:42] <Lee Howard> (applause in the room for v4eit WG)
[07:44:29] <Joel Jaeggli> address utilization efficiency due to fixes assignment of port resources.
[07:46:04] <Joel Jaeggli> we documented in the draft/ compression works pretty well for between 10/100 users per draft.
[07:46:26] <Joel Jaeggli> fred - we'll put of the question until the ADs come back to us
[07:46:34] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Next presentation -----
[07:46:51] <Joel Jaeggli> http://www.ietf.org/proceedings/83/slides/slides-83-v6ops-2.ppt
[07:46:57] <Joel Jaeggli> gang chen
[07:47:02] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Nat64 operational experiences
[07:47:29] <Joel Jaeggli> powerpoint fail
[07:47:43] <iljitsch> should use keynote anyway
[07:47:59] <iljitsch> and maybe a new displayport - VGA adapter
[07:48:29] <iljitsch> is the presentation available in pdf?
[07:48:52] rbarnes joins the room
[07:48:58] <rbarnes> who is presenting?
[07:49:06] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Gang Chen
[07:49:08] <Joel Jaeggli> gang chen
[07:49:13] <rbarnes> thx
[07:51:44] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> slides-83-v6ops-2.ppt <http://www.ietf.org/proceedings/83/slides/slides-83-v6ops-2.ppt>
[07:52:05] <iljitsch> no pdf version?
[07:52:19] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Can't find one
[07:52:24] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> And no slide numbers
[07:52:32] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> I think this is slide #4
[07:52:39] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> "Received Comments"
[07:52:41] <Joel Jaeggli> presentation crasshes ppt on mac
[07:52:52] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> But can be imported in KeyNote
[07:53:01] <Joel Jaeggli> slide 4
[07:53:10] <Joel Jaeggli> yeah it works in openoffice too
[07:53:20] <Joel Jaeggli> which is what I just did
[07:53:40] <iljitsch> what MTU statement?
[07:53:48] <iljitsch> this discussion is entirely at the meta level
[07:54:07] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> It's about comments to the draft
[07:54:22] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> http://tools.ietf.org/html/draft-chen-v6ops-nat64-experience-01
[07:54:47] Répondeur joins the room
[07:54:50] <Maoke Chen> same question on the MTU statement.
[07:55:30] <iljitsch> so the IPv4 servers should set their MTU to at least 1260?
[07:55:30] kawashimam leaves the room
[07:55:30] kawashimam joins the room
[07:55:32] <iljitsch> I said that years ago.
[07:55:40] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> "3.6. MTU Consideration
As compared to the MTU consideration in NAT64-CGN, MTU of IPv4
network are strongly recommeded to set more than 1260. "
[07:55:42] <iljitsch> good work item for v4exit wg
[07:55:59] <iljitsch> oej: where is that from?
[07:56:19] <Joel Jaeggli> comments?
[07:56:23] <Maoke Chen> Maoke Chen from FreeBit, quick question to the presenter: regarding both NAT64-CGN and NAT64-CE, is there the use case in your practice where double NAT64 translation happens?
[07:56:30] <Joel Jaeggli> remi D -
[07:56:45] <Joel Jaeggli> support as a wg document
[07:56:56] <iljitsch> fibrib: how would that work? You can only get from 6 to 4 once
[07:57:03] <iljitsch> or there also needs to be a nat46
[07:57:08] <Joel Jaeggli> fred - question would it be appropiate to a v4exit wg?
[07:57:17] <Joel Jaeggli> I have not objection to v6ops
[07:57:19] Ole Troan joins the room
[07:57:26] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> With both nat64 and nat46 in the same network we do need loop detection ;-)
[07:57:53] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Lee Howard on mike
[07:57:57] <iljitsch> first someone needs to come up with a working nat46, though :-)
[07:58:19] <Répondeur> I hope that v4exit stays v4exit, not v4-life-support.
[07:58:32] <Joel Jaeggli> lee howard - do you need to know if there is no v4 exit
[07:58:38] <Maoke Chen> iljitsch: i have the same feeling but somewhere else there is a discussion about that so i ask the presenter about the practical experiences.
[07:58:45] <iljitsch> I think v4 life support has been going on for the last hour...
[07:58:47] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> v4survival will be an under-the-radar group
[07:59:03] <Joel Jaeggli> fred -assuming there is no v4 exit is this a document suitable for this wg
[07:59:07] <Joel Jaeggli> no humming
[07:59:14] <Joel Jaeggli> no humming
[07:59:16] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Bob Hinden on mike
[08:00:05] <Joel Jaeggli> hold off decisions for something that may happen in two… lets just deal with it now
[08:00:18] <Joel Jaeggli> land of Sarte...
[08:00:21] <Joel Jaeggli> no-exit
[08:00:24] nemo joins the room
[08:00:34] <Lee Howard> Hell is other working groups
[08:01:01] <Joel Jaeggli> lees draft - hum in favor of bringing it into this wg
[08:01:08] <Joel Jaeggli> minor hum opposed
[08:01:14] <Lee Howard> Donley's draft
[08:01:16] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Next presentation -------
[08:01:22] <Joel Jaeggli> sorry yes
[08:01:27] <Répondeur> Chairs: *if* there is time, I'm happy to go ahead and give the draft-townsley-troan-ce-transitioning talk today. It actually might help to get this issue on the table now, so that potential resolution can be worked on before Thursday when the authors present 6204-bis.
[08:01:35] <Lee Howard> You can have my draft(s) too, if you want.
[08:01:42] <Joel Jaeggli> http://www.ietf.org/proceedings/83/slides/slides-83-v6ops-3.ppt
[08:01:44] <Répondeur> (sigh, "Reponder" is Mark Townsley - need to fix my adium settings)
[08:01:47] <Joel Jaeggli> nah we already have them
[08:02:02] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Mark - us /nick
[08:02:08] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> use /nick to change name
[08:02:39] <iljitsch> /nick this is not IRC
[08:02:58] Répondeur leaves the room
[08:03:00] Olle E. Johansson, Paris, France for the IETF 83 (GMT +1) is now known as OEJ
[08:03:10] <OEJ> Even if it's not IRC it works for me
[08:03:24] wmtownsley joins the room
[08:03:27] <york> /nick Dan York
[08:03:29] <wmtownsley> test
[08:03:33] OEJ is now known as Olle E. Johansson
[08:03:38] Lee Howard leaves the room
[08:03:51] <york> Doesn't seem to work using "Messages" on a Mac :-)
[08:04:02] <york> I'm not using Adium
[08:04:34] Suzanne joins the room
[08:04:35] <Olle E. Johansson> "Easy to deploy and available today" - Marketing talk at IETF :-)
[08:05:17] blourdel joins the room
[08:05:47] <Olle E. Johansson> Slide #4
[08:05:51] <Joel Jaeggli> is deployed...
[08:06:08] <Joel Jaeggli> easy or good for you is another matter
[08:06:40] iljitsch leaves the room
[08:06:44] iljitsch joins the room
[08:08:00] <Olle E. Johansson> Anyone that knows an Android platform that has no IPv6. The first speaker mentioned they exist. (Got question on twitter)
[08:08:10] wmtownsley leaves the room
[08:08:13] <Olle E. Johansson> Slide #5 - Cost effective transition to IPv6
[08:08:20] <Olle E. Johansson> Page #6 Timeline of draft
[08:09:28] <Olle E. Johansson> Slide #7 - Versions of the draft changes
[08:09:33] <Olle E. Johansson> Or something like that
[08:09:41] <Olle E. Johansson> Actually just "changes"
[08:09:47] Lee Howard joins the room
[08:09:51] Andrew McGregor leaves the room
[08:10:09] weiyinxing joins the room
[08:10:21] <Joel Jaeggli> android platform with no v6? or one with no v4?
[08:10:39] <york> It was in the WIDE testing draft
[08:10:50] <Olle E. Johansson> Android With no v6 :-)
[08:11:10] <york> There was a comment that at least one Android device that locked out users with v6.
[08:12:05] <Olle E. Johansson> Who's talking on the mike?
[08:12:08] <Joel Jaeggli> remi d -
[08:12:12] <Lee Howard> Remi Depres
[08:13:15] <Joel Jaeggli> lorenzo - like this as a hack to get carrier pigeon quality of v4 onto a v6 only devices.
[08:13:38] <Joel Jaeggli> don't understand why the device needs a /96
[08:13:46] <Joel Jaeggli> should work with a /128
[08:14:17] <Joel Jaeggli> fred - I'd like to walk through the history of this draft on the list.
[08:14:32] <Joel Jaeggli> saw it as wg shopping
[08:14:53] <Joel Jaeggli> asked softwire and behave what they though of it.
[08:15:23] <Joel Jaeggli> about 20 responses yes
[08:16:22] wmtownsley joins the room
[08:16:35] <Joel Jaeggli> the right resolution to the normative language problem is not to remove the normative language, but split doucment into bcp for providing the service, the other draft is the report on the service. seperate those two things.
[08:16:49] weiyinxing leaves the room
[08:18:04] <Joel Jaeggli> lorenzo - it's not a bcp right
[08:18:21] <Joel Jaeggli> this is one thing you can do using these existing technology
[08:19:05] mem joins the room
[08:19:48] <Joel Jaeggli> philip mathews - service providers like standard
[08:19:49] iljitsch leaves the room
[08:20:23] <Joel Jaeggli> mark t - going to echo most of what lorenzo said.
[08:20:29] iljitsch joins the room
[08:20:37] <Joel Jaeggli> barely passes my thresholds
[08:21:17] <Janos Mohacsi> I very much against BCP prefer informational
[08:21:29] <Joel Jaeggli> what I worry about when these things get coupled together
[08:22:18] <Joel Jaeggli> remi d - support informational, think it requires coordination with softwire
[08:23:14] <Joel Jaeggli> lorenzo - all in favor of activities that do not require coordination with softwire. weakening this draft if necessary to not require that would be ok.
[08:23:58] <Joel Jaeggli> hui-deng - useful as short term before 3gpp release 8
[08:24:12] <Joel Jaeggli> lorenzo - remove that behavior and call it done.
[08:24:46] <Olle E. Johansson> "a carrier-pigeon quality service"
[08:28:56] <Joel Jaeggli> test on /96 removal removing objections to progress
[08:29:02] <Joel Jaeggli> about 6 hands
[08:29:30] <Joel Jaeggli> 3 or so would be ameliorated by this
[08:31:37] ryuji@CA joins the room
[08:31:49] <Joel Jaeggli> m dec cisco - need comments on list address
[08:31:49] <wmtownsley> Chairs (on earlier agenda bash): My homenet co-chair just reminded me that we are expected in the rtg-area meeting on Thursday which clashes with v6ops. So, it would be really nice if I could fit in draft-townsley-troan-ce-transitioning today...
[08:32:50] <Joel Jaeggli> I don't think we can make that happen
[08:33:25] <Olle E. Johansson> Next presentation -------
[08:33:27] <Joel Jaeggli> given the outstanding issues with 6204bis
[08:33:39] <Olle E. Johansson> Stateless Source address mapping for ICMPv6 packets
[08:33:42] <york> http://tools.ietf.org/html/draft-ietf-v6ops-ivi-icmp-address
[08:33:48] ray joins the room
[08:33:51] <Joel Jaeggli> http://www.ietf.org/proceedings/83/slides/slides-83-v6ops-5.ppt
[08:34:13] ray leaves the room
[08:34:19] <iljitsch> please people, PDF versions!
[08:34:20] <wmtownsley> I actually think that talking about the ce-transitioning draft today would help reach a resolution for 6204-bis by thursday.
[08:34:35] <iljitsch> I really don't want to fire up Powerpoint or some such just to read a few slides
[08:35:29] <wmtownsley> At least we would have a WG read on the topic, and a few days to work on it.
[08:35:39] geir leaves the room
[08:36:23] <iljitsch> with all the v4 space in use loose uRPF is dead anyway
[08:36:26] <Olle E. Johansson> PDFs with page/slide numbers !!!
[08:36:35] <iljitsch> so I don't see the too big using RFC 1918 issue
[08:36:54] <Olle E. Johansson> We have a page number behind the head of the chair
[08:36:55] <iljitsch> oej: many PDF readers provide these
[08:36:58] <Olle E. Johansson> Page #5
[08:38:53] <Joel Jaeggli> mark - yeah i agree kinda, but we're low on minutes
[08:39:02] <Olle E. Johansson> Page #6 - RFC 5837 issue
[08:39:35] <Joel Jaeggli> yeah, I really don't want to recive packets from this address
[08:42:12] <Joel Jaeggli> comments
[08:42:49] <Joel Jaeggli> fred - what do want?
[08:43:04] <Joel Jaeggli> we'll take it out to wg last call on the mailing list
[08:43:37] <Joel Jaeggli> brian carpenter - how should we proceed with icp guidance.
[08:44:32] <Olle E. Johansson> Lee Howard on mike
[08:44:44] <Olle E. Johansson> Dan York waiting for mike
[08:44:50] <Joel Jaeggli> lee howard - it's a good document
[08:45:04] <Joel Jaeggli> dan york - I think it's very useful guidance.
[08:45:14] <Joel Jaeggli> joel - don't like it
[08:45:37] <Joel Jaeggli> erik k - it's probably a good starting point.
[08:46:18] <Joel Jaeggli> hum - some yes, no opposition
[08:46:30] <Joel Jaeggli> repost as draft ietf.
[08:46:43] <Olle E. Johansson> next presentation ------------
[08:46:54] <Joel Jaeggli> http://www.ietf.org/proceedings/83/slides/slides-83-v6ops-8.ppt
[08:47:00] <Olle E. Johansson> IPv6 Flow label for server load balancing - update (speaker sheng Jiang)
[08:47:10] <Olle E. Johansson> page #2
[08:49:05] <iljitsch> why is load balancing in routers out of scope? That's the most important place to do it.
[08:49:27] <Olle E. Johansson> page #4
[08:49:28] <Joel Jaeggli> indeed
[08:49:35] <Olle E. Johansson> And why only http?
[08:50:14] <Joel Jaeggli> l3/l4 ecmp hash however is stateless and they're looking for stickyness
[08:50:42] <Joel Jaeggli> from the label
[08:50:51] <york> The draft mentions use of flow labels for other protocols - http://tools.ietf.org/html/draft-carpenter-v6ops-label-balance-02#section-3
[08:52:08] <iljitsch> it's dangerous to assume that the flow label is never going to change mid-session
[08:52:55] markushx joins the room
[08:53:40] <Lee Howard> can anyone remember what the remainder of the agenda is for this session, given that it changed on the fly?
[08:56:10] <Joel Jaeggli> lorenzo - if the lb trusts the client to determine lb then you probably you.
[08:56:32] <Joel Jaeggli> the draft should state that you don't trust the end host flow label.
[08:56:55] <Joel Jaeggli> you have to do the same amount of work on every single packet.
[08:57:05] markushx leaves the room
[08:57:13] <Olle E. Johansson> We've been on Page #5 for quite a while now...
[09:01:13] Olle E. Johansson leaves the room
[09:01:49] <Joel Jaeggli> who is at the mic
[09:03:00] iljitsch leaves the room
[09:03:48] mem leaves the room
[09:05:28] <Lee Howard> so, no idea on remaining agenda today?
[09:07:13] Olle E. Johansson, Paris, France for the IETF 83 (GMT +1) joins the room
[09:09:22] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Should we continue the presentation and continue discussion afterwards? We're on slide #5 of 9
[09:11:04] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Simon Perrault on mike
[09:11:05] woolf joins the room
[09:11:15] <Joel Jaeggli> who is it the mic
[09:11:19] kbransom leaves the room
[09:11:19] Dominik Elsbroek leaves the room
[09:11:19] rbarnes leaves the room
[09:11:19] Ole Troan leaves the room
[09:11:19] Joel Jaeggli leaves the room
[09:11:22] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Changing flow label in OpenBSD seems like a three line diff
[09:11:33] ryuji@CA leaves the room
[09:11:39] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Erik Kline google on mike
[09:11:47] S73654-TX0FB16B913 leaves the room
[09:12:11] Maoke Chen leaves the room
[09:12:18] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> What happened with the MUC room bridge? People are kicked left and right.
[09:12:24] Ole Troan joins the room
[09:12:53] markushx joins the room
[09:13:06] Maoke Chen joins the room
[09:13:09] Dominik Elsbroek joins the room
[09:13:28] ryuji@CA joins the room
[09:13:40] woolf leaves the room
[09:13:56] kbransom joins the room
[09:14:12] S73654-TX0FB16B913 joins the room
[09:14:45] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> We're now on page 6!!!
[09:16:01] Joel Jaeggli joins the room
[09:16:29] S73654-TX0FB16B913 leaves the room
[09:16:59] ryuji@CA leaves the room
[09:17:21] rbarnes joins the room
[09:17:59] tomek joins the room
[09:18:09] staffan@kerker.se joins the room
[09:18:14] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Page 7 - security considerations was covered
[09:18:35] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Who's talking?
[09:18:48] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Erik from Google now
[09:19:07] iljitsch joins the room
[09:19:11] <iljitsch> if they change source addresses while talking to the same server (between different sessions of course) then obviously things aren't that stable.
[09:20:25] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Joel talking
[09:20:34] <iljitsch> overwriting the flow label: why not write your label somewhere that you actually get to change, such as an MPLS label or something in the ethernet header or some such
[09:20:48] <iljitsch> I really don't see how the upsides here are worth the downsides
[09:20:55] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> http cookies?
[09:21:02] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> ;-)
[09:21:28] nemo leaves the room
[09:21:52] <Joel Jaeggli> erik kline -
[09:23:08] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> Page 9 - THanks
[09:24:22] Janos Mohacsi leaves the room
[09:24:24] <Joel Jaeggli> adrychenko - you can take thesee20 bits and use them for whatever you want
[09:24:39] wmtownsley leaves the room
[09:24:40] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> …"within your own AS"
[09:24:53] <Joel Jaeggli> yeah
[09:24:54] Janos Mohacsi joins the room
[09:25:11] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> "This draft is fairly controversial in this wg" - Fred
[09:25:11] jpc leaves the room
[09:25:11] Janos Mohacsi leaves the room
[09:25:44] jpc joins the room
[09:25:58] jpc leaves the room
[09:27:15] york leaves the room
[09:27:21] tomek leaves the room
[09:27:22] <Joel Jaeggli> hum, inditerminate hum yes or no for question of is it a wg documnet candidate or not
[09:27:25] <Olle E. Johansson, Paris, France for the IETF 83 (GMT +1)> That concludes the meeting
[09:27:28] Atarashi Yoshifumi leaves the room
[09:27:28] Ole Troan leaves the room
[09:27:32] Olle E. Johansson, Paris, France for the IETF 83 (GMT +1) leaves the room
[09:27:34] Dominik Elsbroek leaves the room
[09:27:34] kawashimam leaves the room
[09:27:36] blourdel leaves the room
[09:27:37] iljitsch leaves the room
[09:27:51] dudisaki leaves the room
[09:27:56] Janos Mohacsi joins the room
[09:28:01] Akira Kato leaves the room
[09:28:12] wmtownsley joins the room
[09:28:15] kbransom leaves the room
[09:29:01] <Joel Jaeggli> there are several point of feedback on the idea of the draft they should be taken individually rather than as consensus
[09:29:14] Janos Mohacsi leaves the room
[09:29:28] wmtownsley leaves the room
[09:29:58] Diego R. Lopez leaves the room
[09:29:58] arifumi leaves the room
[09:30:18] jpc joins the room
[09:31:24] rbarnes leaves the room
[09:32:40] staffan@kerker.se leaves the room
[09:33:09] Bjoern A. Zeeb leaves the room
[09:33:44] Wes George leaves the room
[09:37:01] Dominik Elsbroek joins the room
[09:38:55] Joel Jaeggli leaves the room
[09:39:28] Lee Howard leaves the room
[09:42:21] jpc leaves the room
[09:44:30] markushx leaves the room
[09:45:03] Maoke Chen leaves the room
[09:47:56] Joel Jaeggli joins the room
[09:52:31] Joel Jaeggli leaves the room
[09:55:11] Joel Jaeggli joins the room
[10:00:41] Dominik Elsbroek leaves the room
[10:11:31] Olle E. Johansson, Paris, France for the IETF 83 (GMT +1) joins the room
[10:11:39] Olle E. Johansson, Paris, France for the IETF 83 (GMT +1) leaves the room
[10:49:28] Dominik Elsbroek joins the room
[10:56:03] Joel Jaeggli leaves the room
[10:57:35] Dominik Elsbroek leaves the room
[10:58:07] Atarashi Yoshifumi joins the room
[10:58:25] Atarashi Yoshifumi leaves the room
[10:59:10] Joel Jaeggli joins the room
[11:01:04] Dominik Elsbroek joins the room
[11:05:48] Diego R. Lopez joins the room
[11:06:43] Diego R. Lopez leaves the room
[11:07:29] tomek joins the room
[11:08:54] tomek leaves the room
[12:08:36] tomek joins the room
[12:18:41] wmtownsley joins the room
[12:24:41] wmtownsley leaves the room
[12:29:53] wmtownsley joins the room
[12:37:45] arifumi joins the room
[12:50:06] Dominik Elsbroek leaves the room
[12:50:34] wmtownsley leaves the room
[12:56:58] Dominik Elsbroek joins the room
[12:58:18] arifumi leaves the room
[13:01:31] Joel Jaeggli leaves the room
[13:05:29] Joel Jaeggli joins the room
[13:11:20] nemo joins the room
[13:11:26] tskj leaves the room
[13:13:02] nemo leaves the room
[13:14:40] Dominik Elsbroek leaves the room
[13:19:18] tomek leaves the room
[13:21:08] tomek joins the room
[13:45:57] Dominik Elsbroek joins the room
[13:46:47] tomek leaves the room
[13:53:38] tomek joins the room
[14:03:38] Dominik Elsbroek leaves the room
[14:05:30] tomek leaves the room
[14:07:41] tomek joins the room
[14:10:55] Joel Jaeggli leaves the room
[14:27:24] Joel Jaeggli joins the room
[14:31:14] Dominik Elsbroek joins the room
[14:31:17] tomek leaves the room
[14:31:33] tomek joins the room
[14:31:40] tomek leaves the room
[14:36:19] Joel Jaeggli leaves the room
[14:42:50] Joel Jaeggli joins the room
[15:15:31] naikumar@cisco.com leaves the room
[15:26:02] Ole Troan joins the room
[15:26:53] Ole Troan leaves the room
[15:28:19] Ole Troan joins the room
[15:46:47] tomek joins the room
[15:59:45] Joel Jaeggli leaves the room
[16:00:09] Joel Jaeggli joins the room
[16:14:49] Ole Troan leaves the room
[16:15:53] Ole Troan joins the room
[16:34:53] tomek leaves the room
[16:35:33] Ole Troan leaves the room
[16:55:56] Joel Jaeggli leaves the room
[17:00:06] Joel Jaeggli joins the room
[17:00:20] wmtownsley joins the room
[17:01:35] wmtownsley leaves the room
[17:03:00] Dominik Elsbroek leaves the room
[17:04:40] wmtownsley joins the room
[17:17:29] Joel Jaeggli leaves the room
[17:27:31] wmtownsley leaves the room
[17:59:33] tomek joins the room
[18:24:28] tomek leaves the room
[20:14:05] Ole Troan joins the room
[20:16:09] Ole Troan leaves the room
[20:18:14] Ole Troan joins the room
[20:28:29] Ole Troan leaves the room
[20:30:50] Ole Troan joins the room
[20:34:35] Ole Troan leaves the room
[21:27:42] Dominik Elsbroek joins the room
[21:31:38] tomek joins the room
[21:34:47] tomek leaves the room: Replaced by new connection
[21:34:48] tomek joins the room
[21:36:40] Dominik Elsbroek leaves the room
[21:37:24] Dominik Elsbroek joins the room
[21:57:47] Dominik Elsbroek leaves the room
[22:49:54] tomek leaves the room
[22:53:52] tomek joins the room
[23:00:09] tomek leaves the room
[23:08:46] Joel Jaeggli joins the room
[23:35:48] wmtownsley joins the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!