IETF
v6ops
v6ops@jabber.ietf.org
Thursday, July 21, 2016< ^ >
Alejandro Acosta has set the subject to: IETF 95 V6OPS Monday Afternoon
Room Configuration
Room Occupants

GMT+0
[11:46:57] Meetecho joins the room
[11:59:01] Mikael Abrahamsson joins the room
[11:59:01] Michael Weiss joins the room
[11:59:02] Alex Aust joins the room
[11:59:02] Tore Anderson joins the room
[11:59:06] Scott Morizot joins the room
[11:59:51] Mikael Abrahamsson (scribe) joins the room
[12:00:12] <Mikael Abrahamsson (scribe)> I'll be jabber scribe.
[12:00:19] Scott Morizot_8452 joins the room
[12:00:25] Scott Morizot leaves the room
[12:00:30] Mikael Abrahamsson leaves the room
[12:00:34] Victor Kuarsingh joins the room
[12:01:27] Lee Howard joins the room
[12:01:47] BillC joins the room
[12:02:17] joel jaeggli joins the room
[12:02:30] joel jaeggli has set the subject to: IETF 96 agenda
[12:03:04] joel jaeggli has set the subject to: v6OPS IETF 96 Thursday afternoon
[12:03:04] <Tore Anderson> I'm here
[12:03:17] Ole Trøan joins the room
[12:03:24] Akira Nakagawa joins the room
[12:03:31] <Tore Anderson> I'm on via Meetecho
[12:04:52] joel jaeggli_4255 joins the room
[12:07:13] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[12:08:44] <Lee Howard> Mikael Abrahamsson at the mike
[12:09:12] Dave Thaler joins the room
[12:09:18] Nathalie Trenaman joins the room
[12:09:22] <Mikael Abrahamsson (scribe)> Mikael Abrahamsson
[12:10:18] honlue joins the room
[12:10:38] <Tore Anderson> can you hear me?
[12:10:45] <Ole Trøan> Not remotely.
[12:10:46] Juan P. Cerezo joins the room
[12:10:53] <Tore Anderson> am I supposed to do something to start the presentation mode?
[12:11:04] <Meetecho> Tore Anderson: did you get in queue?
[12:11:04] <Tore Anderson> there is a "presentation" button I'm pressing
[12:11:08] <Lee Howard> Fred is showing slides,
[12:11:17] <Meetecho> we were not aware of any remote presentation here today
[12:11:24] <Meetecho> if you want to speak
[12:11:27] <Meetecho> you need to get in queue
[12:11:29] <Meetecho> hand icon
[12:11:32] <Meetecho> next to your name
[12:11:33] Simon Pietro Romano joins the room
[12:11:48] <joel jaeggli> it sounds like you logged in as an obsorver
[12:12:01] <Tore Anderson> yes
[12:12:03] <Meetecho> clicking the hand icon will change him to presenter
[12:12:16] <Tore Anderson> I did
[12:12:25] becarpenter joins the room
[12:12:30] <Meetecho> (y)
[12:12:33] furry13 joins the room
[12:12:39] <Lee Howard> Meetecho, you guys are the best!
[12:12:47] <Meetecho> :)
[12:12:56] dudisaki joins the room
[12:14:15] <Ole Trøan> Let's give him 2002::16. ;-)
[12:14:23] <Ole Trøan> /16
[12:16:26] <Mikael Abrahamsson (scribe)> Dave Thaler
[12:18:28] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[12:20:05] Benno Overeinder joins the room
[12:20:14] nygren joins the room
[12:20:35] furry13 leaves the room
[12:21:04] <Dave Thaler> @Ole: ha.
[12:21:45] Juan P. Cerezo leaves the room
[12:24:14] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[12:24:55] <Mikael Abrahamsson (scribe)> David Schnazi
[12:25:57] Audric Schiltknecht joins the room
[12:28:03] <Mikael Abrahamsson (scribe)> Jen Lenkova
[12:29:20] <Mikael Abrahamsson (scribe)> Pierre Pfister
[12:29:46] furry13 joins the room
[12:30:48] <Mikael Abrahamsson (scribe)> Dmitry Kohmanyk
[12:31:21] <honlue> Musa Stephen HONLUE: why a /16, and not a /24 or /28 for instance? I think a /16 is too vast.
[12:32:07] <Lee Howard> If everyone got their own prefix from their RIR for every transition technology they used, how much space would it be?
[12:32:24] evyncke joins the room
[12:32:29] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[12:33:02] <Mikael Abrahamsson (scribe)> honlue: Several other people have made the same point, do you still want your name on +1 for that, at the mic?
[12:33:25] <Mikael Abrahamsson (scribe)> David Lamparter speaking
[12:33:54] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[12:34:08] Benno Overeinder leaves the room
[12:34:30] <Mikael Abrahamsson (scribe)> David Schnazi
[12:35:34] Michael Weiss leaves the room
[12:36:25] <Mikael Abrahamsson (scribe)> Dave Thaler
[12:37:38] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[12:38:05] <Mikael Abrahamsson (scribe)> Joel Jaeggli
[12:38:51] <Mikael Abrahamsson (scribe)> Mohamed Boucadair
[12:40:48] <Mikael Abrahamsson (scribe)> David Schnazi
[12:41:01] geir joins the room
[12:41:43] geir leaves the room
[12:42:32] Caciano Machado joins the room
[12:45:16] <Mikael Abrahamsson (scribe)> Mikael Abrahamsson (me) asked for that clarification from David
[12:45:55] <Mikael Abrahamsson (scribe)> David Schnazi
[12:47:20] <Mikael Abrahamsson (scribe)> Brian Carpenter
[12:48:41] furry13 leaves the room
[12:49:36] Benno Overeinder joins the room
[12:50:17] Tore Anderson_8662 joins the room
[12:50:35] Tore Anderson leaves the room
[12:54:23] <Tore Anderson_8662> Meetecho: The slides displayed in the web interface are clipped on the left/right sides
[12:55:00] <Meetecho> Tore Anderson_8662: glitch in capture card, cropping should be minimal though
[12:55:04] Matthew Sargent joins the room
[12:55:10] <Mikael Abrahamsson (scribe)> meetecho: what's the best course of action, should we unplug and plugin again the vga cable going to the capture card?
[12:55:20] <Meetecho> Mikael Abrahamsson (scribe): nope, better not to touch anything
[12:55:30] <Tore Anderson_8662> Meetecho: ack, I'll be fine, just thought I'd let you know
[12:55:35] <Meetecho> (y)
[12:57:13] <Ole Trøan> 16:9 vs 4:3?
[13:00:11] Alex Aust leaves the room
[13:00:12] <Lee Howard> Mikael Abrahamsson
[13:00:16] <Meetecho> 16:9 are usually fine
[13:00:27] <Caciano Machado> https://www.ietf.org/proceedings/96/slides/slides-96-v6ops-1.pdf
[13:00:37] <Mikael Abrahamsson (scribe)> Dave Thaler
[13:00:56] <Ole Trøan> Rule5.5 is RA SA + PIO
[13:01:15] <Mikael Abrahamsson (scribe)> ole, so never RIO?
[13:01:28] <Ole Trøan> https://datatracker.ietf.org/doc/draft-ietf-6man-multi-homed-host/
[13:01:40] <Dave Thaler> RIO is what gives you a next-hop for a given prefix
[13:01:42] Jorge I. Blanco joins the room
[13:01:42] <Ole Trøan> Never RIO
[13:01:44] <Dave Thaler> (destination prefix)
[13:02:03] <Ole Trøan> RIO is destination. PIO is source.
[13:02:16] <Ole Trøan> Check out the 6man draft also
[13:02:17] <Mikael Abrahamsson (scribe)> "advertised prefix" is kind of vague.
[13:02:31] <Mikael Abrahamsson (scribe)> but I guess the RFCs are more exact in language.
[13:03:01] <Dave Thaler> RFC 6724 is pretty precise.  so to get the full logic of 5.5, both PIO and RIO (or default route) are relevant
[13:03:03] <Ole Trøan> of course RIO influences next-hop selection too.
[13:03:11] <Ole Trøan> so it is all interconnected.
[13:03:14] <Dave Thaler> right
[13:03:22] joel jaeggli leaves the room
[13:03:48] <Mikael Abrahamsson (scribe)> I've always wondered how this applies to multicast as well, because I have problems in Linux that I never can figure out why it sends an MLD join out a certain interface.
[13:03:50] <Ole Trøan> And we have also had proposals on an SRIO option
[13:04:14] <Dave Thaler> @Mikael, the multicast join apis have the app choose the interface
[13:04:24] <Ole Trøan> @Mikael: Let's just deprecate MLD for link-local multicast and you don't have to wonder any more. :-)
[13:04:57] <Mikael Abrahamsson (scribe)> ole: this was for MLDv2 S,G join to GUA S and G address.
[13:05:05] <Ole Trøan> Ack
[13:05:26] <Ole Trøan> To this slide: Host just participates in a SADR enabled RP?
[13:05:45] <Dave Thaler> "RP"?
[13:05:53] <Ole Trøan> Sorry. Routing protocol
[13:06:39] <Mikael Abrahamsson (scribe)> ole, I don't think the idea is that there is any routing protocol the host participates in.
[13:06:44] <Dave Thaler> right
[13:06:47] <Mikael Abrahamsson (scribe)> the host only has RAs to go on
[13:06:52] <Dave Thaler> no changes to host
[13:07:12] <Mikael Abrahamsson (scribe)> dave, well, it will require some changes... or clarifications.
[13:07:28] <Mikael Abrahamsson (scribe)> hosts generally don't do this today
[13:07:36] <Dave Thaler> this is saying a router can have 2 addresses and send an RA from each one, in order to use rule 5.5
[13:07:36] <Ole Trøan> Well, given that such a large portion of hosts are actually routers (e.g. hosts with tethering or VMs) perhaps we should rethink that.
[13:08:16] <Dave Thaler> so far I like this proposal
[13:08:27] <Ole Trøan> @Dave Yes, and this scenario only matters if the host is directly connected to the two exit routers right? If it is deeper down, then you'd expect the router to do SADR.
[13:08:30] <Mikael Abrahamsson (scribe)> dave, same here, it seems to do most of what I've beeb babbling about the past few years.
[13:09:14] <Ole Trøan> What if red prefix flaps at high frequency?
[13:09:16] <Dave Thaler> this is showing how a router can do SADR and make hosts do the right thing without requiring host changes
[13:09:35] <Ole Trøan> I really dislike the coupling of address lifetimes with reachability
[13:09:37] <Dave Thaler> does not require host being directly connected to more than one router.
[13:09:44] <Dave Thaler> there canbe 1 router with 2 addresses
[13:09:46] <Ole Trøan> And typically the network has a very restricted view of "red prefix goes down"
[13:09:58] <Ole Trøan> @Dave Yeah, noticed that.
[13:10:06] <Mikael Abrahamsson (scribe)> Jan Zorz
[13:10:07] <Ole Trøan> This is to avoid SADR enabled router then?
[13:10:19] <Lee Howard> Jan Zorz
[13:10:28] <Lee Howard> Žorž
[13:10:29] <Dave Thaler> no, it's a way of SADR-enabling a router to avoid changing hosts
[13:10:45] <Ole Trøan> But if router does SADR, then there is no need for this.
[13:10:58] <Dave Thaler> yes there is
[13:11:07] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[13:11:23] <Ole Trøan> @Dave As a way to signal reachability you mean?
[13:11:34] <Dave Thaler> this is just how to influence which source address the host chooses for a given destination, or (if oyu want) for the default
[13:11:42] Benno Overeinder leaves the room
[13:12:08] <Mikael Abrahamsson (scribe)> Chris Bowers
[13:12:20] <Ole Trøan> Right, I would claim that the application / transport / session layer has to figure that out for itself... because adding all these moving parts only helps for local network reachability...
[13:12:35] <Ole Trøan> E.g. a typical site edge router doesn't participate in global routing.
[13:13:20] <Ole Trøan> The only thing that can detect end to end reachability is the application... so having the network solve a corner case of the bigger problem ....
[13:13:24] <Mikael Abrahamsson (scribe)> David Lamparter
[13:13:41] <Dave Thaler> it's saying that R3 can send 2 RA's to the host as if it were like R1+R2 in the same box
[13:14:02] <Ole Trøan> Yes, but the only reason to do that is to convey reachability information, right?
[13:14:23] Juan P. Cerezo joins the room
[13:14:38] <Dave Thaler> well, not the only reason.  using 2 RAs can also leverage rule 5.5 in hosts that support that, since the next hops will be different in the two RAs
[13:14:40] <Mikael Abrahamsson (scribe)> Pierre Pfister
[13:14:42] <Ole Trøan> And conveying routing information in address lifetimes... well, I have some architectural issues with that.
[13:15:21] <Ole Trøan> Dave, if the router does SADR, it doesn't make any difference, the router forwards based on host's choice of SA anyway...
[13:15:38] <Dave Thaler> but the difference is how the host chooses the SA
[13:15:43] <Ole Trøan> the two RAs could convey different policy information I suppose which could trigger the host to make a different choice, but that's all.
[13:15:46] <Ole Trøan> Right.
[13:15:48] Benno Overeinder joins the room
[13:16:12] <Dave Thaler> so it makes a difference since you can better influence how the host picks the SA
[13:16:29] <Ole Trøan> But based on what information? The network doesn't know about end to end path behaviour...
[13:16:36] <Dave Thaler> (which isn't just about reachability, it might also be for policy/TE)
[13:17:03] <Ole Trøan> Wouldn't the host (and the end user) be better off trying all possible paths and sticking with those that work
[13:17:28] <Ole Trøan> E.g. networks is trying to influence the host to pick a source that doesn't work.
[13:17:36] joel jaeggli joins the room
[13:17:36] <Dave Thaler> often yes, but different apps have different needs and do different things.
[13:18:39] Jorge I. Blanco_7145 joins the room
[13:18:45] <Dave Thaler> many choose one path per destinatoin address (not every address pair), since that's what RFC 3484 did
[13:19:28] joel jaeggli leaves the room
[13:19:28] <Dave Thaler> Windows changed to use address pairs (not just one per dest addr) in around windows 7 i think.
[13:19:35] Lee Howard joins the room
[13:19:47] <Ole Trøan> Right, regardless to get host based multi-homing to work and to fulfill the requirements in the multi-homing requirements RFC, it needs to be a lot more clever than it is today. at least something like MPTCP
[13:19:48] <Dave Thaler> and of course apps can use their own algorithm if they want
[13:20:08] Lee Howard leaves the room
[13:20:10] <Ole Trøan> I would like applications to behave like MOSH.
[13:20:26] <Ole Trøan> But we're far out of the no host changes space. ;-)
[13:21:03] <Ole Trøan> Which I think is required to support MHMP anyway... I've run a MHMP network at home for a few years, until I disabled the second link. We're far away from making this workable.
[13:21:11] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[13:24:11] <Mikael Abrahamsson (scribe)> Brian Carpenter
[13:25:57] <Mikael Abrahamsson (scribe)> Chris Bowers
[13:26:35] <Ole Trøan> +1 to Carpenter
[13:26:55] <Jorge I. Blanco_7145> :|
[13:27:31] <Ole Trøan> PI space solves that.
[13:28:03] <Ole Trøan> MPTCP solves it
[13:28:20] Akira Nakagawa leaves the room
[13:28:24] Akira Nakagawa joins the room
[13:28:51] Benno Overeinder leaves the room
[13:28:52] <evyncke> PI solves a lot of issues :-)
[13:29:13] <Mikael Abrahamsson (scribe)> Pierre Pfister
[13:29:16] <Mikael Abrahamsson (scribe)> PI doesn't scale.
[13:29:56] <Ole Trøan> @Mikael: Which do you prefer, doesn't work or doesn't scale? ;-)
[13:30:02] Akira Nakagawa leaves the room
[13:30:05] Akira Nakagawa joins the room
[13:31:13] <Mikael Abrahamsson (scribe)> Ole: I prefer solutions like MPTCP.
[13:31:38] Lee Howard joins the room
[13:32:06] Lee Howard leaves the room
[13:32:23] <Lee Howard> Anyone using Jabber over ietf-nat64 or ietf-v6ONLY?
[13:33:51] <Mikael Abrahamsson (scribe)> Dave Thaler
[13:34:26] Randy Turner joins the room
[13:35:24] Benno Overeinder joins the room
[13:35:32] Victor Kuarsingh leaves the room: Disconnected: closed
[13:36:15] <Mikael Abrahamsson (scribe)> Benedikt Stockebrand
[13:36:47] Carlos M. Martinez joins the room
[13:36:52] <Ole Trøan> The sad fact is that today IPv6 NAT works the best for these cases. But please don't relay that to the mike. :-)
[13:37:05] Paolo Saviano joins the room
[13:37:10] <Dave Thaler> :)
[13:37:37] <Lee Howard> It's okay, we Noted your comment Well
[13:37:44] <Mikael Abrahamsson (scribe)> "6MAN CHAIR: NATS ARE GREAT!"
[13:38:05] <Ole Trøan> :-)
[13:38:19] <Mikael Abrahamsson (scribe)> David Lamparter
[13:38:21] Marco Ragonai joins the room
[13:38:42] Francis Dupont joins the room
[13:38:57] Francis Dupont leaves the room
[13:39:57] <Mikael Abrahamsson (scribe)> Tim Chown
[13:41:34] <becarpenter> Lorenzo again
[13:41:35] <Mikael Abrahamsson (scribe)> Lorenzo Colitti
[13:43:29] <Mikael Abrahamsson (scribe)> Pierre Pfister
[13:45:11] <Mikael Abrahamsson (scribe)> Philipp Tiesel
[13:46:05] <Mikael Abrahamsson (scribe)> Chris Bowers
[13:47:10] <Mikael Abrahamsson (scribe)> David Lamparter
[13:49:32] <Mikael Abrahamsson (scribe)> Barbara Stark
[13:49:52] nygren leaves the room: Disconnected: Replaced by new connection
[13:49:52] nygren joins the room
[13:51:35] <Mikael Abrahamsson (scribe)> Geoff Huston
[13:53:15] Paolo Saviano leaves the room
[13:53:34] Benno Overeinder leaves the room
[13:54:47] furry13 joins the room
[13:56:04] BillC joins the room
[13:56:09] BillC leaves the room
[13:57:46] <furry13> I’m on the Jabber from NAT64 SSID
[13:57:53] joel jaeggli joins the room
[13:57:56] <furry13> (not sure who asked about it earlier)
[13:57:59] <furry13> works like a charm ;)
[13:58:06] Benno Overeinder joins the room
[13:58:19] <Lee Howard> Was me. Cool, good to hear.
[13:59:43] <nygren > I'm also on Jabber from the NAT64 SSID.  (I sometimes need to switch back since my VPN and DNS64 client doesn't play well together for some things.)
[13:59:46] joel jaeggli leaves the room
[14:00:04] <nygren > are slides from this posted?  not linked from agenda page.
[14:00:28] <furry13> I think they are presented from an  USB stick, so not yet most likely..
[14:00:51] <furry13> The similar slide deck was presented at IEPG on Sunday and the slides are uploaded already
[14:01:11] <Lee Howard> Yeah, this presentation was a suggestion during agenda bashing. So, no.
[14:01:20] <furry13> http://iepg.org/2016-07-16-ietf96/ipv6-survey-july-2016.pdf
[14:01:42] <furry13> (IEPG slides, I think it is very similar but maybe not exactly the same)
[14:02:01] becarpenter leaves the room: offline
[14:02:35] <Mikael Abrahamsson (scribe)> Tobias Mayer
[14:03:39] furry13 leaves the room
[14:03:51] dudisaki leaves the room
[14:04:02] nygren leaves the room
[14:04:11] Benno Overeinder leaves the room
[14:04:18] Tore Anderson_8662 leaves the room
[14:04:21] Audric Schiltknecht leaves the room: La machine se met en veille
[14:04:30] evyncke leaves the room
[14:04:37] Caciano Machado leaves the room
[14:04:37] Jorge I. Blanco_7145 leaves the room
[14:04:38] Randy Turner leaves the room
[14:04:38] Akira Nakagawa leaves the room
[14:04:39] BillC leaves the room
[14:04:41] Simon Pietro Romano leaves the room
[14:05:19] honlue leaves the room
[14:05:27] joel jaeggli joins the room
[14:05:56] Meetecho leaves the room
[14:06:32] Lee Howard leaves the room
[14:06:46] Mikael Abrahamsson (scribe) leaves the room
[14:06:46] Scott Morizot_8452 leaves the room
[14:07:24] joel jaeggli leaves the room
[14:08:17] honlue joins the room
[14:09:34] joel jaeggli joins the room
[14:10:47] honlue leaves the room
[14:11:32] joel jaeggli leaves the room
[14:14:04] joel jaeggli joins the room
[14:17:20] Carlos M. Martinez leaves the room
[14:20:08] BillC joins the room
[14:22:48] Dave Thaler leaves the room
[14:24:12] honlue joins the room
[14:25:02] honlue joins the room
[14:25:54] honlue leaves the room
[14:26:31] Barbara Stark joins the room
[14:26:37] Barbara Stark leaves the room
[14:27:28] Lee Howard joins the room
[14:58:08] Carlos M. Martinez joins the room
[14:58:39] BillC leaves the room
[15:08:29] BillC joins the room
[15:13:51] honlue leaves the room
[15:17:23] honlue joins the room
[15:19:39] BillC leaves the room
[15:23:21] BillC joins the room
[15:38:44] dudisaki joins the room
[15:40:03] Lee Howard leaves the room
[15:42:41] honlue leaves the room
[15:48:40] BillC leaves the room
[15:54:27] furry13 joins the room
[16:17:04] dudisaki leaves the room
[16:21:20] furry13 leaves the room
[16:32:08] furry13 joins the room
[16:38:55] dudisaki joins the room
[16:39:15] dudisaki leaves the room
[16:56:29] nygren joins the room
[17:00:30] nygren leaves the room
[17:20:55] furry13 leaves the room
[17:27:02] Carlos M. Martinez leaves the room
[17:38:08] joel jaeggli leaves the room
[17:40:13] joel jaeggli joins the room
[17:42:11] joel jaeggli leaves the room
[17:46:07] joel jaeggli joins the room
[17:49:49] joel jaeggli leaves the room
[17:52:10] joel jaeggli joins the room
[17:54:18] joel jaeggli leaves the room
[18:06:55] joel jaeggli joins the room
[18:08:45] joel jaeggli leaves the room
[18:15:09] joel jaeggli joins the room
[18:17:48] joel jaeggli leaves the room
[18:23:21] joel jaeggli joins the room
[18:24:44] joel jaeggli leaves the room
[18:34:36] joel jaeggli joins the room
[18:36:27] joel jaeggli leaves the room
[18:46:04] joel jaeggli joins the room
[18:48:39] joel jaeggli leaves the room
[18:51:18] joel jaeggli joins the room
[18:55:01] joel jaeggli leaves the room
[20:07:07] BillC joins the room
[20:16:10] BillC leaves the room
[20:45:27] joel jaeggli joins the room
[20:47:16] joel jaeggli leaves the room
[21:11:48] joel jaeggli joins the room
[22:09:25] joel jaeggli leaves the room
[22:20:28] Benno Overeinder joins the room
[22:20:47] Benno Overeinder leaves the room
[22:42:30] Lee Howard joins the room
[22:42:37] furry13 joins the room
[22:50:51] Lee Howard leaves the room
[23:12:06] furry13 leaves the room
Powered by ejabberd - robust, scalable and extensible XMPP server Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!