IETF
v6ops
v6ops@jabber.ietf.org
Thursday, 17 November 2011< ^ >
Patrik Halfar has set the subject to: v6ops -- IETF80
Room Configuration

GMT+0
[02:15:21] Bing Liu leaves the room: Replaced by new connection
[02:15:22] Bing Liu joins the room
[02:17:15] Bing Liu leaves the room
[02:17:32] Bing Liu joins the room
[02:50:15] Bing Liu leaves the room
[04:40:54] Joel Jaeggli joins the room
[04:51:43] becarpenter joins the room
[04:52:09] mcr joins the room
[04:54:20] TACHIBANA toshio joins the room
[04:56:52] Chris Griffiths joins the room
[05:00:26] Yuri Nawata joins the room
[05:01:35] genki_yasutaka joins the room
[05:01:41] Warren Harrop joins the room
[05:01:55] <Joel Jaeggli> our agenda is commencing
[05:02:31] <Joel Jaeggli> bing liu - Analysis and recommendation for the ULA usage <http://tools.ietf.org/html/draft-liu-v6ops-ula-usage-analysis>
[05:02:37] furry joins the room
[05:02:51] <Joel Jaeggli> incorporates feedback from original feedback
[05:04:32] <Joel Jaeggli> initial draft that is
[05:04:44] dudisaki joins the room
[05:05:00] frnkblk joins the room
[05:07:16] kawashimam joins the room
[05:10:29] Arturo Servin Ü joins the room
[05:11:27] <mcr> mic: draft-ietf-ipv6-ula-central-02 rather old now. I wish it wasn't. But I think ULA-C needs more thought for enterprise use.
[05:12:28] <Joel Jaeggli> I'll take to mic in the question phase
[05:12:40] Bjoern A. Zeeb joins the room
[05:12:40] <mcr> ty.
[05:12:41] behcet.sarikaya joins the room
[05:13:12] Dave Thaler joins the room
[05:13:22] tatsuji joins the room
[05:13:46] <mcr> "ULA-only"?
[05:14:22] <Joel Jaeggli> fred b - you'd work better if you stepped back an inch or two
[05:15:01] <Joel Jaeggli> devices like cars or airplanes or buildings are connected to the internet all the time
[05:15:24] <Joel Jaeggli> the reason I got a class was I need nsome prefixes for my lab bench
[05:16:03] <Joel Jaeggli> ula if you're not going ot connect then a ula is easy and painless
[05:16:27] <Joel Jaeggli> rfc 6296 largest usage uses public prefixes
[05:16:40] <Joel Jaeggli> so there's nothing special of using a ula.
[05:16:55] <Joel Jaeggli> I'd be hesitant to make a strong recomendation
[05:17:02] <Joel Jaeggli> a ula is an option
[05:17:23] Atarashi Yoshifumi joins the room
[05:17:55] <Joel Jaeggli> the better security argument I'm sorry we've been over it so many times it doesn't do that.
[05:18:24] Arturo Servin Ü leaves the room
[05:18:39] <Joel Jaeggli> benifit of renumbering we got a comment from the iesg that a ule could be a benifit. I couldn't find one.
[05:20:08] <Joel Jaeggli> I drew a picture amazingly like that and what I had in mind was that 200 or so different companies that are treated as though they're part of it.
[05:20:32] <Joel Jaeggli> you could take one /48 and use that across a static nat boundry
[05:21:02] <Joel Jaeggli> busine privte routing actually does make sense.
[05:21:28] <Bjoern A. Zeeb> fast forward please
[05:21:29] <mcr> mic: there is a term for business/business private routing, which is COIN. ULA-R fails, because there is no whois. ULA-C would win.
[05:22:38] <Joel Jaeggli> arfumi - in 6man we're trying to better handle ula combined with ipv6 global
[05:24:09] HUI joins the room
[05:24:11] <Joel Jaeggli> lee howard - don't want to go into a dicussion about whether nat is good, your positon is inconsistent
[05:24:55] <Joel Jaeggli> seichi k - I consult for enterprise, lot of questions about how do we use ula, providing some guidance, we should dicuss how not to use it.
[05:25:08] <Joel Jaeggli> it's just another 1918 but that's not true
[05:25:43] <Joel Jaeggli> ? - take the lanaguage about nat out.
[05:25:47] Arturo Servin Ü joins the room
[05:25:51] <furry> I should say that ridiculously large number of packets blocked on my edge are due to ULA and even site local source..
[05:25:59] <Joel Jaeggli> the arguement that we shouldn't run multiple prefixes is a bad arguement
[05:26:33] <Joel Jaeggli> documents that specif where it makes sense is a smaller set of cases.
[05:27:30] <Joel Jaeggli> erik kline - instead of removing it, should say should not
[05:27:49] <Joel Jaeggli> ? - could apply to our use case in home control networks.
[05:28:09] <Joel Jaeggli> these nodes are very contrained and very stable
[05:29:02] <Joel Jaeggli> slide 9 as defined in 6lowpan we can create short addresses e.g. with /64
[05:29:18] <Joel Jaeggli> time chown slide 6 or 7 6 lowpan case.
[05:29:55] zcao joins the room
[05:29:56] danwing joins the room
[05:30:09] <Joel Jaeggli> mark t - +1 to tim
[05:30:31] <Joel Jaeggli> presentation in quebec that microsoft and apple did - hey guys this is what we want
[05:31:05] <Joel Jaeggli> offered to put together a seperate document as opposed to homenet.
[05:31:23] <Joel Jaeggli> fred can we get a hum on whether that document would be useful.
[05:31:53] <Joel Jaeggli> brian c - you mean a scenario document?
[05:31:59] arifumi joins the room
[05:32:22] <Joel Jaeggli> mark t- impacation how the devices expect the network to behave.
[05:33:14] hoyaj@jabber.org joins the room
[05:34:25] <Joel Jaeggli> fred - maybe it belongs coming from the iab.
[05:34:26] danwing leaves the room
[05:34:42] <Joel Jaeggli> lorenzo - does such a document belong here?
[05:34:47] vincent.levigneron joins the room
[05:35:11] <Joel Jaeggli> would it be this is what I learned or what I think it should do.
[05:35:17] danwing joins the room
[05:35:20] <Joel Jaeggli> ok I'll sit down.
[05:36:08] <Joel Jaeggli> lorenzo - I thinks that's written into 3484
[05:36:42] <Joel Jaeggli> jari a - as the resident ula skeptic, ran a few network in none of those cases do we use ulas
[05:37:21] <Joel Jaeggli> I don't like the vision of ula being an manadatory component in the network.
[05:37:41] <Joel Jaeggli> bob - the scenarios are in the ula document.
[05:38:12] <Joel Jaeggli> the thing about nat is it's very seperate from the disucssion of what address you use.
[05:38:28] <Joel Jaeggli> ulas have a role
[05:39:34] <Joel Jaeggli> ? - the comments about nat and security which I support. the can be globally routed but we don't
[05:40:08] <Joel Jaeggli> becaue we don't want the garbage , if you lekak them it's not fine.
[05:40:48] <Joel Jaeggli> tom h - without ula we have no local connectivity - that use case is necessary
[05:41:25] <Joel Jaeggli> brian c - ulas used with nptv6 cat is out of the bag
[05:41:56] <Joel Jaeggli> to bob a lot of scenarios are descibed but that hasn't been enough.\
[05:42:16] Ole Troan joins the room
[05:42:20] Bill joins the room
[05:42:43] tsavo_work@jabber.org/Meebo joins the room
[05:42:59] fdupont joins the room
[05:44:00] <Joel Jaeggli> jari a - local connectivity is obviously very much needed, have to deal with the question of status of your connection versus the lease time on the prefix you have.
[05:44:13] redaka joins the room
[05:44:22] Colman Ho joins the room
[05:44:27] Colman Ho leaves the room
[05:44:39] Colman Ho joins the room
[05:45:07] <Joel Jaeggli> ted leman - a typical home gateway doesn't know, ulas get you over th hump, a prefix you get from your provider you are supposed to give up when it expires
[05:45:47] <behcet.sarikaya> ted lemon
[05:46:09] <Joel Jaeggli> yeah
[05:47:22] <Joel Jaeggli> brian carpenter - icp guidance.
[05:48:48] Warren Harrop leaves the room
[05:56:52] <Joel Jaeggli> lorenzo - I object to the word encourage
[05:57:07] <Joel Jaeggli> can we can make it more attractive to do v6
[05:57:45] <Joel Jaeggli> they'll move when they find it necessary so I don't think we need to encourage them.
[05:58:25] <Joel Jaeggli> fred - so tell me the isp is going to translate for me?
[05:58:29] <Joel Jaeggli> nope
[05:59:29] <Joel Jaeggli> I know that an inside out approach would not have worked for us
[06:01:43] <Joel Jaeggli> joel - can't imagine doing it anywhere else.
[06:02:06] <Joel Jaeggli> lorenzo - whitelisting draft was stuck for a long time it will cease to be relevant soon.
[06:03:10] <Bjoern A. Zeeb> Next time just have one slot and be a bit faster;)
[06:03:42] <Joel Jaeggli> tim chown - enterprise equivalent I'd be happy to working on
[06:05:21] <Joel Jaeggli> benidict hiledbrad - get an isp, you have to fix all sorts of stuff, it can be a major overhaul can be difficult ot assess. services instrumentation.
[06:06:31] hoyaj@jabber.org leaves the room
[06:07:23] <Joel Jaeggli> ? from china telecom - transition technology may be required?
[06:08:16] <Joel Jaeggli> mark t - it's only been recently that outside-in has become popular
[06:08:57] <zcao> we need time control, anyway...
[06:11:12] <Joel Jaeggli> hui deng - how to handle dos attack / load balancer
[06:11:33] <Joel Jaeggli> brain c - look at the draft send comments.
[06:12:11] <Joel Jaeggli> http://tools.ietf.org/agenda/82/slides/v6ops-1.pdf
[06:13:09] Francisco Arias joins the room
[06:13:38] Francisco Arias leaves the room
[06:25:31] <Joel Jaeggli> alain d - this is a useful approach - worth documenting, the discussion we had in int area about identifying each user
[06:26:04] <Joel Jaeggli> fred b - your comments are well taken
[06:26:46] <Joel Jaeggli> what she's trying to show them is v6 traffic.
[06:27:32] <Joel Jaeggli> lorenzo - who are we trying to convince ?
[06:27:45] naito.kengo@jabber.org joins the room
[06:28:07] <Joel Jaeggli> if I were a content provider I wouldn't want to run this in production.
[06:31:04] <Joel Jaeggli> lorenzo - there is no hope of making that mapping scale
[06:32:43] <Joel Jaeggli> hui deng - how to track those subscribers - they can not use to location platform.
[06:33:48] <Joel Jaeggli> lorenzo - I would heistate to put the seal of approval from vsops
[06:34:11] <Joel Jaeggli> lorenzo - online legal intercept doesn't work.
[06:34:38] <Joel Jaeggli> use a load balancer or application layer proxy.
[06:34:40] furry leaves the room
[06:35:12] <Joel Jaeggli> ted lemon - the fact that lorenzo is so concerned about geolocation .
[06:35:31] <Joel Jaeggli> most egregios problem is abuse detection.
[06:35:51] <Joel Jaeggli> not everybody is google.
[06:36:00] <Joel Jaeggli> - alain d
[06:37:39] <Joel Jaeggli> dan w - conflates nat 64 in front of user and nat64 int front of a datacenter are completely different
[06:38:38] <Joel Jaeggli> zhen co http://tools.ietf.org/agenda/82/slides/v6ops-5.pdf
[06:45:17] Ole Troan leaves the room
[06:49:44] <Joel Jaeggli> dan wing - are you recomending that hosts that are ipv6 enabled use ipv6. in their fqdn
[06:50:10] <Joel Jaeggli> you don't do that.
[06:51:12] <Joel Jaeggli> jari - we shouldn't discuss that nor should the draft discuss it
[06:52:29] <Joel Jaeggli> lee howard - the slides provide a lot more detail than the draft
[06:53:20] <Joel Jaeggli> confused who the audience for this document.
[06:54:42] <Joel Jaeggli> hui deng - i belive the people reading this draft will come from the 6to4 rfc.
[06:54:58] <Joel Jaeggli> mak t - sunsetting 6rd
[06:56:17] <Joel Jaeggli> strives to allow 6rd virtual and physical interface to remain seperate.
[06:58:31] tsavo_work@jabber.org/Meebo leaves the room
[07:01:24] zcao leaves the room
[07:01:54] danwing leaves the room
[07:03:41] <Joel Jaeggli> fred - ask for it to be here due to requirements imposed on 6204bis
[07:04:02] <Joel Jaeggli> alain d - tink this will have a home in softwire
[07:04:30] <Joel Jaeggli> lorenzo - in 6304 we don't define how they can exist at the same time.
[07:04:38] <Joel Jaeggli> user want's to prefer native.
[07:04:43] behcet.sarikaya leaves the room
[07:04:58] <Joel Jaeggli> operator wants to limitraffic on the br.
[07:05:47] <Joel Jaeggli> no way for the ce to know that native v6 exists.
[07:06:22] Arturo Servin Ü leaves the room
[07:06:31] TACHIBANA toshio leaves the room
[07:07:32] Atarashi Yoshifumi leaves the room
[07:07:42] kawashimam leaves the room
[07:07:44] fdupont leaves the room: Computer went to sleep
[07:07:45] Chris Griffiths leaves the room
[07:08:06] HUI leaves the room
[07:08:20] Joel Jaeggli leaves the room
[07:08:31] arifumi leaves the room
[07:10:01] Bjoern A. Zeeb leaves the room
[07:10:55] danwing joins the room
[07:12:01] Colman Ho leaves the room
[07:12:31] Bill leaves the room
[07:14:38] Joel Jaeggli joins the room
[07:14:52] Arturo Servin Ü joins the room
[07:15:00] Arturo Servin Ü leaves the room
[07:16:08] TACHIBANA toshio joins the room
[07:16:51] danwing leaves the room
[07:17:02] vincent.levigneron leaves the room
[07:20:02] dudisaki leaves the room
[07:20:02] becarpenter leaves the room
[07:21:02] redaka leaves the room
[07:22:02] Dave Thaler leaves the room
[07:28:02] naito.kengo@jabber.org leaves the room
[07:28:48] Colman Ho joins the room
[07:28:57] Colman Ho leaves the room
[07:31:19] TACHIBANA toshio leaves the room
[07:35:02] arifumi joins the room
[07:35:30] naito.kengo@jabber.org joins the room
[07:41:02] tatsuji leaves the room
[07:46:43] arifumi leaves the room
[07:47:50] naito.kengo@jabber.org leaves the room
[07:55:05] Bjoern A. Zeeb joins the room
[08:01:40] Geoff Huston joins the room
[08:02:04] Geoff Huston leaves the room
[08:09:47] Bjoern A. Zeeb leaves the room
[09:27:04] Joel Jaeggli leaves the room
[09:28:21] genki_yasutaka leaves the room
[09:29:43] Joel Jaeggli joins the room
[10:43:04] Joel Jaeggli leaves the room
[10:45:49] vincent.levigneron joins the room
[10:46:16] vincent.levigneron leaves the room
[10:47:05] vincent.levigneron joins the room
[10:52:25] Joel Jaeggli joins the room
[10:59:16] vincent.levigneron leaves the room
[11:19:02] Joel Jaeggli leaves the room
[12:15:28] mcr leaves the room
[12:17:50] vincent.levigneron joins the room
[13:12:37] frnkblk leaves the room
[13:31:44] Joel Jaeggli joins the room
[14:12:18] vincent.levigneron leaves the room
[14:13:08] vincent.levigneron joins the room
[16:02:19] vincent.levigneron leaves the room
[16:07:19] vincent.levigneron joins the room
[18:06:21] vincent.levigneron leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!