From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============2296651469799019601==" MIME-Version: 1.0 From: Christoph Paasch To: mptcp at lists.01.org Subject: Re: [MPTCP] Call for ideas for a presentation about MPTCP Upstream project at NetDev 0x12 in July Date: Mon, 30 Apr 2018 09:25:01 -0700 Message-ID: <20180430162501.GO19260@MacBook-Pro-6.local> In-Reply-To: alpine.OSX.2.21.1804300809340.4562@seskerx-mobl.amr.corp.intel.com X-Status: X-Keywords: X-UID: 570 --===============2296651469799019601== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable On 30/04/18 - 08:37:01, Mat Martineau wrote: > = > On Sun, 29 Apr 2018, Matthieu Baerts wrote: > = > > Hi Mat, > > = > > Thank you for your review and input! > > = > > On Sat, Apr 28, 2018 at 2:44 AM, Mat Martineau wrote: > > = > > Hi Matthieu - > > = > > On Fri, 27 Apr 2018, Matthieu Baerts wrote: > > = > > Hello, > > = > > Following yesterday's discussion about NetDev 0x12, here is= a proposition of mail to send to NetDev committee. I > > already put some comments. Note that I have followed the su= bmission guidelines from: > > https://www.netdevconf.org/0x12/submit-proposal.html#propos= als > > = > > Please comment it before Monday morning. Sorry for the rush= but the deadline is the 1st of May :) > > = > > * Name(s) of the submitter(s): Christoph Paasch (Appl= e), Mat Martineau > > = > > (Intel), Peter Krystad (Intel), Ossama Othman (Intel) and m= yself, Matthieu Baerts (Tessares) > > = > > I wrote down the names of people who participated in the di= scussion in the ML and during the weekly meetings. I can > > add more people if more people would like to join the prepa= ration and presentation of this tutorial. > > @Christoph, Mat, Peter, Ossama: can I write your names ther= e? > > = > > = > > Ossama won't be able to attend. It looks like most sessions have = 1 or 2 presenters, so I don't know if 4 is too many. I'm sure > > they'll give us some feedback if they want to limit the number of= presenters. > > = > > = > > OK thank you! > > Yes indeed, I am sure they will say something if we are too many. > > = > > * Title of the submission: MPTCP: from the basic to a= n upstreamable base > > = > > = > > We certainly need a better title, please comment! > > = > > = > > "Multipath TCP: Present Use Cases and an Upstream Future" ? > > = > > = > > I was trying to find something that catch the attention but it is maybe= not needed for these kind of presentation. > > = > > > * Label (one of moonshot, nuts'n bolts, hands-on): hands-= on > > = > > "hands-on" seems to be the correct one according to the Sub= mission Types. > > https://www.netdevconf.org/0x12/submit-proposal.html#types > > = > > = > > It's a little bit of "nuts & bolts" too, but I agree that hands-o= n is the better match. > > = > > = > > Thank you! > > = > > * Submission type (one of talk, presentation, worksho= p): tutorial, > > = > > instructor-led sessions (minimum 1 hour long and not to exc= eed 1.5 hours. The instructor will go over the technology > > either through code review or execution and interact with t= he attendees.) > > = > > I guess there is a typo here: presentation should be replac= ed by tutorial in the guidelines. > > = > > Estimate of length of time for presentation: 1h > > = > > = > > That's what we agreed yesterday but I can change. > > = > > Affiliations of submitters (needed for conflict of in= terest check): Apple, > > = > > Intel, Tessares > > = > > Description of proposal: > > =C2=A0 =C2=A0 A project to add an implementation of t= he MultiPath TCP protocol to the > > = > > Linux kernel is in progress by a small community. The goal = of this tutorial is to discover what is this TCP extension > > (RFC 6824), what are the different use-cases already in pro= duction by some companies and what are the challenges to > > upstream MPTCP. We hope having interactive discussions and = getting feedback from experienced developers will help us > > in this task of easily bringing MPTCP to all Linux users, a= technology already used by millions of people. > > = > > = > > =C2=A0 =C2=A0 In a bit more detail, we will start wit= h a basic introduction of MPTCP. > > = > > A few use-cases will be presented with a demo to explain ho= w useful this protocol is in today's Internet and how it > > can be extended with API's like Netlink and BPF. > > = > > = > > I recognize how Netlink is associated with a userspace path manag= er, but what's the BPF extensibility you're referring to? > > = > > = > > It was only to mention that it would be possible to extend MPTCP with e= BPF. > > Even if it is not already available in the current Open-Source implemen= tation, there is already the possibility to get a version with a > > programmable scheduler: https://progmp.net > = > If we don't have something specific to our effort with BPF, I'm not sure = it > belongs in the summary. But I don't have a strong opinion about it. I agree that we don't need BPF in the summary here. We can talk about its potential during the presentation though. But, here in the summary it is a bit more distracting IMO. > = > > = > > Then we will have some explanations about how MPTCP is curr= ently implemented. This current implementation is quite > > intrusive and that is certainly not something we would like= to have upstream. We would like to express what we have > > in mind to change that, with some samples and initiate disc= ussions. > > = > > = > > My edit of the above: > > = > > """ > > = > > A community project is underway to add Multipath TCP to the upstr= eam Linux kernel. This tutorial will introduce the audience to > > this TCP extension (RFC 6824), show some use cases already in pro= duction, and discuss the challenges in converging on an upstream > > MPTCP implementation. > > = > > = > > Should we insist on the "discussion" part? Like saying: "discuss with t= he audience"? "have interactive discussions"? > > It is maybe not needed but it was mainly to express the interactivity, = it is a tutorial, not just a "simple" presentation. > = > I think it's ok to mention that - while I left the word "discuss" in there > for that reason, maybe that's not obvious or clear enough. I think that "discuss with the audience" or "have interactive discussions" makes it more clear what we want to achieve. Christoph > > We will use the current MPTCP implementation to demonstrate the u= tility of the protocol on today's internet, and to show how this > > implementation can currently be extended with netlink and BPF. Th= is not only has practical application for deploying MPTCP now, > > but also illustrates how the APIs and code will need to evolve in= order to properly coexist with the optimized Linux TCP core we > > all rely on. We will discuss our ideas for bringing MPTCP to the = upstream kernel so the technology is available to all Linux > > users. > > = > > = > > To "attract" people, should we mention that the current implementation = is already used by millions of users? > > I like how your improve the last bit :-) > = > Sure. I liked that in your text, but ran out of time figuring out how to = fit > it in to the last sentence :) > = > > = > > """ > > = > > Feel free to edit/merge/expand/discard as needed :) > > = > > = > > Thank you for your edit, it is indeed cleaner! > > = > > =C2=A0 =C2=A0 For more information about this project: > > = > > https://github.com/multipath-tcp/mptcp_net-next/wiki > > = > > = > > I guess I can keep this, right? > = > Yes, that was my intent. > = > = > Thanks, > = > Mat > = > = > > = > > Please feel free to comment this as well. We are still far = from the max 350 words limit we found last time. But on > > the other hand, I can no longer find this limit on their we= bsite :) > > = > > = > > Hope the above is helpful. Thanks again for your work on this pro= posal. > > = > > = > > Yes it is, thank you for your help! > > = > > Matthieu > > = > > = > > = > > Mat > > = > > = > > = > > Thank you for your help! > > = > > Have a good day/evening, > > Matthieu > > = > > On 25/04/2018 21:58, Mat Martineau wrote: > > = > > Hi Matthieu - > > = > > On Fri, 20 Apr 2018, Matthieu Baerts wrote: > > = > > Hello, > > = > > = > > NetDev 0x12 is coming to Montr=C3=A9al this sum= mer: July 11th to 13th, 2018. > > = > > = > > We already talked about this event on this ML a= nd at our weekly meetings but here is a > > summary of the discussions we had: > > = > > =C2=A0- we would like to have a presentation th= ere mainly to get feedback and advice from other > > kernel developers > > =C2=A0- a presentation would clearly indicate t= hat this MPTCP Upstream project exists and we could > > get help from more developers > > =C2=A0- we would like to indicate that having M= PTCP upstream is asked by different companies, some > > are even ready to contribute ; it is then impor= tant to have MPTCP upstream > > = > > = > > Also note that: > > = > > =C2=A0- David Miller will not be present in Mon= tr=C3=A9al [1] but other main contributors should be > > there (we don't have a list) > > = > > = > > Side note, in the past day David reiterated his state= ment about not attending or supporting the > > conference: > > = > > https://marc.info/?l=3Dlinux-netdev&m=3D1524668272033= 01&w=3D2 > > = > > =C2=A0- A presentation by Octavian Purdila abou= t "MPTCP Upstreaming" has already been given in > > 2015 (NetDev 0.1) [2] > > =C2=A0- 3 types of presentation are available: = talks, tutorials and workshops [3] > > =C2=A0- Call for Presentation Proposals closes = on May 1st, 2018. > > = > > = > > The current idea we briefly discussed during ou= r weekly meetings is to give a tutorial: > > = > > =C2=A0- It is not useful to give almost the sam= e presentation as the one of Octavian > > =C2=A0- It will allow us more flexibility someh= ow to explain what is MPTCP, the different > > use-cases, why it is important to have it upstr= eam and what problems we are currently facing. > > =C2=A0- David Miller and many other kernel deve= lopers will go to LPC in November: a good place to > > give a talk this time. > > = > > = > > Do you have any ideas on what we could show in = this tutorial? > > = > > = > > I recently discussed with my colleague Olivier = Bonaventure who has a lot of experiences in > > giving different introductions and more about M= PTCP and here is what he suggests: > > = > > =C2=A0- A first part about a basic introduction= of MPTCP > > =C2=A0- Indicate different use-cases -- if poss= ible with a "closed demo" to be sure it is working > > -- asking people to setup something is not easy= in 1h, max 1h30. > > = > > = > > =C2=A0From the description at [3], either "instructor= -led" 60-90 minute tutorials or "student-participation" > > 2-3 hour sessions are possible. The closed demo maps = well to their "instructor-led" category. Looking at > > the schedule, the past two Netdev Conferences have ha= d one tutorial each, of 60-70 minutes. I think it > > helps to be closer to an hour in length to hold the a= udience's attention. > > = > > =C2=A0- Then trying to have interactive discuss= ions or explanations about how MPTCP is currently > > implemented or should be implemented if it goes= upstream, e.g.: for MPTCP, we need to have > > extra TCP Options, we need to support middlebox= es, we need to link subflows of the same > > connection together, we need a scheduler, a PM,= etc. > > = > > = > > It's difficult to predict how interactive an audience= will be. I've only attended one Netdev Conference, > > and it seemed like there were a lot more people with = expertise and interest in drivers, lower layers > > (XDP, BPF, TC, netfilter), and network topology/simul= ation. Discussion around middlebox support and the > > userspace API might have more audience interaction. I= f we want to drive a discussion, we could try to > > strike a balance between topics for the broader audie= nce and those with more knowledge of TCP internals. > > (Hopefully some TCP internals people are still planni= ng to attend) > > = > > =C2=A0- Of course, we should focus our discussi= ons on the upstreaming aspect, e.g. reducing the > > footprint of MPTCP in the current TCP stack: wh= at are we allow to do, what not. It is linked > > to many previous discussions we had on this ML,= e.g. why we need more indirect function calls > > and how to reduce the impact, etc. > > = > > = > > The previous talk ([2]) had a section like this. I ha= ven't watched it recently, I should look at it again > > to see what kind of questions the audience was asking= . As you mentioned above we should be careful to > > have new content compared to the previous session. > > = > > =C2=A0- If we have time, we could discuss about= how users could interact with MPTCP: enable it per > > connection, control the path manager, maybe the= scheduler, etc. > > = > > = > > What do you think about this? Feel free to comm= ent and even propose completely different > > ideas! > > = > > = > > Thank you for outlining these ideas. I see that this = topic is on our meeting agenda so it will be good to > > discuss the tutorial there. > > = > > = > > [1] https://lists.01.org/pipermail/mptcp/2018-M= arch/000379.html > > [2] https://www.youtube.com/watch?v=3Dwftz2cU5S= Zs > > [3] https://www.netdevconf.org/0x12/submit-prop= osal.html > = > -- > Mat Martineau > Intel OTC --===============2296651469799019601==--