Internet Engineering Task Force 94, 2 Nov 2015

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: IETF94-PALS-20151102-1300.webm

Description

PALS meeting session at IETF94
2015/11/02 1300

A

Thanks yep, so I think it's that's right. You can't let them know. You have her bring your back here and get me home, so they surely go yeah yeah.

B

A

About Carrie thought about very natural, Ginny and I could occur. It music here looks like we're certain. Oh they saw me is my man. It's just that well, I'm not prepared to travel, yet it out yes, but they actually.

A

A

B

A

Deal come on, yeah do with you the hope. Oh yeah, wow, yeah, Libby's, going to sky yeah, the guys really being a nice time, but absolutely and then we can pour my heart yeah.

A

Cancel me it's a cattle on it, which I make 10 5 yeah yeah yeah, but it missed it.

A

They were looking hey folks.

C

It's it's one minute after a one o'clock now, but we're going to wait a couple more minutes because it is right after lunch and it's a it's a short agenda so weak. So it's not like we're really cramped for time. So we're going to start the blue sheets going around. So please sign the blue sheets and we're actually will start the agenda at about five past. One.

A

It's funny because I got some facial allowing.

A

That's all done so so I have to make this last well well and when I got.

D

A

It was top of my madness. Oh yeah,.

A

D

A

Late 2013, that's about my last video here, I'm duper.

A

So it's to your own, petty, but but I still works. Fine in the official company image is with, it is windows 7 and they gave me. You know, windows, 7 p.m. that all the country tools and it has the the corporate tunnel. Okay thing is because, while ways pc only right but everything works in am running, see these turkeys, Justine, so lots of fiber based encryption, I'm using connect your check it out. That's what their.

A

A

So you actually have to run is gonna. Give it one more time. On the back years, we have to run to the Super to get design fee a disk encryption I, the video we made. Those yet didn't do that for you for what they did not do. That.

B

A

Yeah, so so all of my corporate stuff is a name isn't one does.

A

Yeah, exactly everything for our crisis. Support died, yeah, whatever actually.

A

Its processes and you in the nasty grams like two days with manatees yeah, we told you about you, sorry.

B

A

A

Yes, those changes and anything else, that's coming from the iesg. They are never the vast that.

C

Was it fine yeah, so no.

A

No just make that change, I think we're done so yeah.

A

Yes, what no, what she do is I would send it to the guy and Deborah, and once he says it looks good, then you up would.

C

A

C

Ok, let's get started.

C

My name is Andy malice, I'm, one of your coach. Here's, the other coach here is Stuart Bryant to, unfortunately, was not able to join us in Yokohama for this meeting. So instead of Stuart joining me here up in front, we have day cynic rope.

C

Who is our working group secretary, who's, who's assisting me up here and normally it would be Dave taking minutes and seeing as dave's up here, Scott Mansfield who's, the fellow standing up over there has an waving, has volunteered to take minutes in dave's place, so we're all set and we'll be monitoring jabber as well.

C

So if there's anyone remote I, don't see anyone remote in the job room right now, but if there's anyone listening remotely or watching us on me, dekho, we we don't have incoming meetup need echo for this room. So if you want to ask any questions, we motely please login to the pals java room and we are monitoring it. So you can ask questions remotely. The other thing is for the minutes, we're using etherpad and the cool thing about etherpad.

C

Is that all of you, if you want, can get onto etherpad while you're sitting, and if you see anything, that's got mrs. or so on your free to to to edit yourself as multiple people can work in etherpad simultaneously. So with that the first thing on the agenda, it's a shorter gender, as I said I'll, be showing you in a minute, is I'll, be going through the working group status slide. So once you move to the next slide Dave. So this is the the note.

C

Well, this is Monday and seeing this Monday, some of you may not have had a chance to read the note well yet so I'm actually going to leave it up for a minute just to give you those you who have not yet seen it a chance to look at it. Basically, what it says is this is a meeting that you're attending anything that you see that you see or do or otherwise participate at.

C

The meeting is really um public information for use by the IETF and and so by being here at the IETF meeting, you really are participating and that cludes hallway conversations and anything else. Okay, so this is the agenda, as I said. It's short, we have two items. This presentation is the first item, then. The second item will be the layer, 2 vpls yang model, and that will be from police beset and one thing I want to notice it. What want to point out is that should be draft shot best mpls l2vpn yang z, Ro 0.

C

We had appointed to an old draft right here. Unfortunately, so that's actually draft da bes bes s, mpls l2vpn, yesh 00, and that has been corrected in the online agenda. So if you look in the online agenda, you'll see that that's correct. It's just incorrect in my slides. So the objective for this presentation is to present the progress in terms of yang modeling for layer tues.

C

This includes both vpls and v pw s and the reason that it's a best working group draft is that includes both vpls and EVP n + and so basically, because it has of a VPN which is bgp based and vpls, which is ldp based, we flipped a coin and best one. So it's in the best working group. Okay, we have no new rc's since last meeting, but we're by this time in at the at the next meeting, a brontosaurus.

C

We expect a bunch and the reason for that is that we've been working really hard on priming the pump. This is actually been keeping your working group cheers very busy and your area director very busy, so I think Deborah needs a big hand because she's been extremely busy working for us.

C

So so thank you, debra, and so we have a three drafts now in the RC editors q, as you can see, pals MSP, w protection, visas, Aoife gal and icc p stp, which originally actually was a peewee three draft, and we didn't bother changing at the PALS because it was in the iesg by the time that we meet that we made the changeover and for those of you who are the authors, if you receive email from the RFC editor, please do your best to respond to the RC editor quickly, so that they can get their work done.

C

Okay and also keeping Deborah busy and the working group cheers as well as well as the authors, because the author is sometimes need to do updates is we have five more drafts that are currently under iesg review? Three of those need revised drafts of for the authors of those three drafts, pe3 mpls-tp mac with drawer- and you doesn't see spe please get to work and get those updates done, and you already know exactly what needs to be updated. So so just go ahead and take care of that.

C

Then we have two more that are being actively reviewed by deborah pals congestion considerations and RC. Four, four, four seven bits now I've got a note on our c4 447 bits, which is kind of thing. As you all know, it was our c4 447, which really got started with one of the very first things we didn't pee with me working group, which is the LDP based targeted signaling for pseudo wires and what we want to do with this update.

C

Not only are we bringing it up to date in terms of including all of the arata and everything else. Is this come up over the years, but we've decided it's time to make it a full internet standard and there's a process from making things internet standard. There's an RC 6410 which sells out that process, and there are a number of different criteria that you have to meet in RF e6410 in order to qualify for internet standard, so we're working with debra to show how the RC meets the criteria in our c64 10.

C

So we can then get it's that it's a full internet standard status approved by the iesg. Now one little wrinkle that's come up. Is that, as you know, RC 4447 and biss are all about targeted, ldp signaling. Well, it turns out that ldp itself is not yet internet standard. It's right!

C

Now it's at the draft standard stage which no longer exists because RC 6410 got rid of the draft standard stage, but that's still where ldp sits, and so one of the things that I've been discussing with lower and the chairs of the MPLS working group is to see, if maybe they can get their rears in gear and get ldp itself brought up to inter next full internet standard.

C

So Debra again will be working with the chairs of the MPLS working group to see what we, what we need to do over in the MPLS working group to bring ldp up to internet standard, because if that doesn't happen, then we'll need what's called a down ref in in 44 47 bits to when we refer to ldp cuz ldp is a normative reference, of course, 4444 seven Bissell so we'll see what happens there, but we'll be watching that with interest and I. Think lowa is planning on discussing this at the mpls session on Wednesday. Okay.

C

Next we have drafted a completed working group last call we have one of those pals vpls pim snooping. This actually passed l2vpn working, good last call, which shows you about how long ago it was. It actually went. The working group last call. The author has told me that he has some specific comments that he's received, that he still needs to resolve with the author of the comments, so there may be some changes coming in the future for this draft and a lot of time has passed and and may have faded a bit people's memory.

C

So we're strongly considering once we have a completed draft rerunning, a pals working group last call for it as well. Okay, then we have drafts in working group last call right now we have none, however, that will change soon. We've been informed by the authors that the two drafts you see up here: mpls-tp om, config and n plus TP sooo, the wire / bi-directional LSP- are ready for work group last call, so Stuart and I will be running those probably starting next week.

C

Okay, then, we have some other drafts that are that are working group drafts in various stages of progress. I asked authors for updates and I got updates on two of them. First of all was pals and PCP dual homing protection and dual humming coordination. Those are from the same groups of authors, the authors report that they refreshed both drafts just before the cutoff prior to this meeting.

C

They do make plans to make updates and the next revision and so they're working on that, but meanwhile, before they make those updates, they would like to solicit review comments from the working group. So I would like to really ask the working group if you haven't had a chance. Please read these two drafts send comments to the authors because they would really appreciate it.

C

Next, we have four more drafts that are just basically working group drafts in progress for each of these I would like to ask if any of the authors just happened to be in the audience if they have an update that led to give the microphone first perils endpoint fast protection. Any the author is out there. Okay, it do you have any comments for the microphone that go up to the mic.

D

At the Wyman's closer for the end upon a fast protection and also for the MSC pong protection, we will update to a newer version of the meeting. Okay.

C

Thank you, okay, thank you and then we have two more pairs: LBP vpls broadcast extension. Do we have an author in the room going once going twice and finally pals seamless, v, CC v again any authors in the room, I'm not hearing any, let's move on and that's it any questions, comments, suggestions before we move on to Patrice.

C

No okay, thank you. Just before Patrice gets up and put up the slides has. Is there anyone in the room who has not yet signed blue sheets every one son? Okay, yes, need blue sheet right. There.

E

Hey guys, um my name is pytheas reset I'm working for Cisco, so I'm here today to present the work activity that we are doing for the young model regarding l3p, n, Andy, VPN, I, guess and actually I just want to recover a queue for the four men shoes draft, if there's no more mpls in the name, actually, okay, so it's really in l2vpn, but we decided to take out the mpls. The service can also be used for VX, nine and so on.

E

Auto end cap and and, as you said, the the draft, a mass change, so I mention you're, going back to version 0.

D

All right: ok, ok,.

E

So this is kind of funny, because the the best work work group asked me to be super fast, and here they are asking me to be super slow so and these lights won't cover any yang detail, because the request was to be fast alright. So let's move on.

C

Don't worry about the ten minutes.

D

E

So, just again for the people who doesn't know, this is pretty much the lease of guys working and with us on the yang model. Tons of people, especially under the l2vpn side, B, it's pretty crowded, but I, think things are doing alright, very good effort and some are working very hard in this thing. Regarding the VPN, it's a subset of that. But again it's you can maybe it's a 60-person and forty percent of those people, so just to give you an idea. This is a very well covered by many different company here.

E

So it's a good team effort um to tell you again how this is the way we are working for the l2vpn. It's it's every week, because it's a lot of work to do. We are focusing on the framework, so every Wednesday we have that meeting the a VPN is every two weeks. This was less to do. I guess there, and what we did also is what is very important. Also is the the approach that we are using.

E

If you look at the current draft that our online, you will see, there's not much in terms of meat right, but our intent was to is really to get the framework up and running. So now, if you telling me I, don't see the MTU inside the bridge, I mean yes, it's normal. Okay doing it don't worry. You're gonna come very soon, but we want to make sure that people likes the framework itself.

E

All right so come configuration data still our still the defocus and we're gonna move very quickly on the most likely on operational data next, so you will see a lot of things happening in the next few months.

D

Next, one: okay.

E

um This is a Manchu yeah. We put inventions name for this graph of me, pls I. Guess we flipped a coin. He won now, which is which is a joke, but um yeah before it was under the pulse and pls. We decided to move it to the best and and also we take out the mpls, because our model will be applicable to any technology, including VX, an encapsulation in Prague. We did explain, we good cover v, pw s. We also explain the concept of an endpoint or virtual port.

E

So if people are still question about about that aspect, you can talk to anybody like you meant you or myself will be glad to help you to explain you what this is this concept, and today this is all about pprs, okay, mention yeah.

F

This is general subject: oh yes, himanshu shah from Siena.

F

So in this discussions about layer to stuff you know, we haven't talked about attachment circuits and we always discuss saying that we will take whatever the mef and I Triple E has decided, and so until now that was a that was a ok but I. Think I've been hearing from other people. Also saying that ok, you know, there's a big piece missing here, which is the hell two aspects of it.

F

So how much we want to take it from there how much we want to put it ourselves and how do we coordinate the whole thing is something we need to add. Some point discuss.

E

The good news we have Robert Wilton here in the room, so we have the right guy he's.

D

E

Our interface gang, sir.

G

Yes, so this is robert wilson, so just wanted to say that I'm already working on a draft in net mod for dressing, vlans and ethernet in terms of classifying traffic to sub interfaces or interfaces and I've in terms of how that binds to this model I was thinking be nicely. The attachment circuit was regarded as an interface, whether that applies for all vendors I, don't know, but I think that's the question that needs to be brought about here. We have to chat, you of life are.

F

You taking all.

G

F

Stuff from my idea, billion amia.

G

Know the the height recovering their models, they're covering 82 @ 1, Q and bridge model and nef- aren't really defining this I. Think in terms of what's been defined in ITF is the underlying technology that you could be used in any F to define circuits.

F

So the so the plan is to take your work and then process it here or know. If there's no, no but.

G

Within that mod I, don't think it actually fits dark in here, because, again, terms of why I'm trying to do is it's more generic to just apply to l2vpn supplies to l-3 as well.

F

Is that is that so this is maybe this is a question for Debra's. Is this a general generic work that applies to all the working group and which working group shuda process that info or how does that work? Because I think that's an important piece? We are carrying the layer, two services over the MPLS Network and then there is so much variations of what is the l2 services right?

F

What is an attachment circuit for all kinds of different if you know for EPL EVP L.

D

We've already, the routing area with the ops have already decided that, if its technology specific, it would be done in the working group that has that expertise. So I.

G

Yes, so I don't think vpm vlans, for example, our technology specific I mean in terms of their not specific to LT VPNs. Is that it's answer your question? Or do you mean why's? He not been done in I Triple E can.

D

G

They they're doing the VLAN bridge model they're, not doing an interface based model current.

F

A

D

I think Helen Chen I think there is a plan to cover l two models in I Triple E. It's just that. Currently it's not complete.

D

Compliment yes, they had experienced a problem with a shortage of yang expertise in and moving slowly.

G

So I think the dot wonky, what we do have a VLAN model, the draft model available for doing Bridget and doing and don't want key bridges. They don't have a model for mapping that two interfaces such so, unless you think they can do work on that as well. So.

D

What I know is that they're soliciting feedback on how to adapt to l2 services? Ok,.

G

My wife chance your flood yeah.

D

F

Yeah, you should talk to her or the people who are doing some work in other standards bodies. You want to make sure that it's an all inclusive or it at least it's not conflicting with their work. Otherwise, we will have a lot of problems. I already.

G

I already am okay.

F

Thanks, that's all.

E

All right, thank you um again also. Just let me know that. Okay, since we came out with the vpls via the sorry, the vp, the Pierce model in Prague, and now we have the VPN s.

E

We made it on purpose to have two distinct model right and- and we can emerge that actually this week so and why we've done this is just to simplify our work, so people can follow what we are doing and but you're gonna see that if you look at the v pw s and the vpls they are so close is going to be very, very trivial for us to merge that together and after that, we're gonna beef, it up we're going to put everything that all the bits and bytes for the the features on it.

A

Yeah, just just one more point on and what we were just talking about, jebra brought up a good point. Perhaps the yang coordination pages should note what's going on in the other venues like I Triple E on VLAN, so that there's no confusion about who's doing what and whether there's any overlap. So if I can maybe I can workbook switching hats, the in coordination hat, um maybe I could talk with a few of you offline. We get some of that recorded on the wiki. Okay.

E

Is that wiki maintained by everyone, because last time, I put my stuff, there was pretty light and I'm not sure if it if it reflect the whole the activity it.

A

Is maintained but.

E

It is maintained by plane, while.

A

Authors: okay,.

E

A

The coordinators all.

D

Right you're aware or not, but at the hackathon on saturday they they actually were doing a Yank analog and it's going to be across all the stos. It will catalog all the different yang models going on, so people are interested in what's going on other SDOs, that's that's! That will be a place for it. Okay,.

A

All right, I, don't want to duplicate that work. Now, yeah.

D

A

Point to that catalog.

C

Deborah, are we hosting that catalog, you know, or is it being hosted elsewhere? Okay, very good.

E

All right, so we can move on I'm, don't need to go through a VPN, even though they are not. We did two balls or we can anyway just quickly yeah, please go by just one. Just gonna go very quickly on this for a VPN, there was nothing presented in in Prague. Now we have the first model out.

E

Unisys is very very light, but again it's a framework and the idea is to connect the dots with the merging model that we're gonna get this week so and then beef up with more the basin buys for the VPN on the feature sign.

E

Also, we did separate the two modules from the ethernet segment manager and the e VPN service itself. That's what for the EV can model next slide. There's also other activities happening right now, where we are trying to. Actually we are trying to come up with an L to yang architecture, informative draft just a case of people as common background or understanding.

E

Sometimes it's just on the on the terminology, there's so many different. Well, you saw the list at the beginning to work. There are so many different people from everywhere. Sometimes we're talking about the same thing just using different terms or just we are like drawing the different pictures, but all the pictures. Men the same so we decided that we're gonna come up with kind of a I, don't know a cookbook or something that's going to help people to understand what we are saying here.

E

Next slides l3v p.m. yeah. You can skip that. Actually, when you talk about it, invest and VX.

A

Lesson question about sorry, that's being done in all three sm, though right.

E

40, no, actually it's two things that threesome is the service model. What we are working with small device model: okay, okay, so service one's gonna use the our model. Also, we've been, then just let you know you can also talking with the open, config folks about our model, so they are right now reviewing what II, Men shoe and myself as here. So so at least we can align them, but these guys are. They told me that they haven't start really looking at it or thinking about L do yet.

E

So we are one step ahead, so hopefully you can join us on the effort. Alright, let's keep moving uh yeah, so VX and that's gonna be I. Some people approach me to start something on that front as well. So if you have any interest, please ping me we'll see where we are Patrice.

C

Did did you mean nvg re there.

E

I mean no I put in more general terms, I don't know yet actually I've been asked about three days before I was coming here. So when I was in the plane doing those lights, I put it down. Yeah.

C

Matthews question yeah.

D

Matthew Bochy, if you're going to start something on the exelon and mva, might once also sort of discuss it with the MV 03 working group as well, because there's some young models floating around in that working group, probably relevant to this yeah.

E

No definitely actually, yes, so anyone who knows something about it, please let me know so: I will I, don't know. What's going on there, honestly, alright next light, ok, yeah! So please go through those those young model.

E

If you don't understand anything about the concepts, the framework, please contact me or anybody else on the list or immense you, it's very important because we can emerge the models very soon and and we want to make sure that people's feels that this is the right things to do right next, one: okay, that's it so this is for best.

E

C

Okay, before we break up, is there anyone who has not yet signed the blue sheets? Okay, we need a blue sheet right over there and pricing needs blue sheet I.

A

Think is that person? No it's not, but but down that way, it.

C

Is not yet and any other questions comments how much you? Yes.

F

Himanshu from so I know, we didn't talk about tho IM stuff, but there was a considerable amount of work done in the past or the absurd of our group about the OEM interworking, and I think that at some point we need to cover that in the in the yang model, there is a if not configurational expect.

F

Maybe at least the state related yang model information needs to be worked out, so maybe mia, maybe at some so this that should be covered in our group here, right right and again, the l2 part of it like CFM and whatever the other layer to OEM stuff. Is there the interworking of that? So that also has to be what they work into this stuff. We have to make a yeah.

C

G are you planning on including up v, CC v as well.

F

Yes, I think that's something that we have to take up. Yes, VCC related parameters and all that yeah, okay,.

C

Great okay, any other comments, questions about anything at all having to do with pals, if not you're, free to to enjoy your afternoon. Thank you all very much.

A

I'll take here, it goes to the desk. Okay,.

B

A

Yeah all right so yeah so.

D

D

F

A

Stewart's very end.

D

Yes, yeah nothing.

C

So this is the acceptable way forward. So Stewart saying number one is the answer. Yeah.

A

So just update the draft so that so that number one is reflected in this craft right, so so so choice, one which is.

C

This one right here I'll leave the draft, as is but ed text describes its constraint. What happens if you have to change.

D

Much how to do about it, I have a yes, so.

A

Just add that I update.

C

A

It's done, yeah.

D

I, just a month law makes you are oh yeah, okay, yes,.

A

Yes, so do you think thank you very okay? Yes, absolutely yeah! That's.

C

D

To do yes, we can pick up attack, yes, yeah. It's.

A

Not on yes, I yep so hope to deuce make that change.

C

And update it yeah.

A

C

Done exactly okay,.

D

C

D

D

C

D

In the medium/large on Thursdays, yeah I know.

A

D

A

It's okay, yeah.

D

D

C

A

Absolutely okay, thanks.

A

B

B

C

A

Routing meeting running area needing.

B

A

B

B

A

B

A

C

A

B

C

Here it is running area working, super running every open meeting, that's 50.

B

C

B

D

Are in the country.

A

You could there's also the.

C

Transport work group is meeting right now next door and.

A

The years ago, some interesting things with the.

C

Transport working group transferred everyone.

B

A

They're they're doing things with armed with.

C

Congestion management for real time services and.

A

And circuit breakers.

C

Yes, we oughtta be.

A

B

C

It was a combination.

C

No, no he's he's continued.