[Iccrg] Re: Iccrg Digest, Vol 81, Issue 8
arjuna.sathiaseelan@gmail.com (Arjuna Sathiaseelan) Sat, 27 October 2012 12:04 UTC
From: arjuna.sathiaseelan@gmail.com
Date: Sat, 27 Oct 2012 12:04:52 +0000
Subject: [Iccrg] Re: Iccrg Digest, Vol 81, Issue 8
In-Reply-To: <508bc04d.ef49b40a.25eb.fffffa61SMTPIN_ADDED@mx.google.com>
References: <508bc04d.ef49b40a.25eb.fffffa61SMTPIN_ADDED@mx.google.com>
Message-ID: <CAPaG1Akg8qycZkZ+Y_9mKBOAR_K_9MDj0kF_c1gN263-=AHj-A@mail.gmail.com>
X-Date: Sat Oct 27 12:04:52 2012
this sounds great actually..probably reduced to 3/4th cwnd on ecn marks? arjuna On Oct 27, 2012 12:06 PM, <iccrg-request@cs.ucl.ac.uk> wrote: > Send Iccrg mailing list submissions to > iccrg@cs.ucl.ac.uk > > To subscribe or unsubscribe via the World Wide Web, visit > http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg > or, via email, send a message with subject or body 'help' to > iccrg-request@cs.ucl.ac.uk > > You can reach the person managing the list at > iccrg-owner@cs.ucl.ac.uk > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Iccrg digest..." > > > Today's Topics: > > 1. Agenda (Michael Welzl) > 2. Why don't we stop treating ECN and loss similarly? (Michael Welzl) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Sat, 27 Oct 2012 10:20:13 +0200 > From: Michael Welzl <michawe@ifi.uio.no> > Subject: [Iccrg] Agenda > To: iccrg list <iccrg@cs.ucl.ac.uk> > Message-ID: <9C0FF8B7-8A75-4599-9980-BBA7B76842F0@ifi.uio.no> > Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes > > Hi, > > There is still space on our agenda - currently, it looks like this: > > > ICCRG agenda, IETF 85, Atlanta, Monday, 5 November, 13:00-15:00, Salon A > > Rong Pan: "A new algorithm for dealing with buffer bloat", 20+10 min > > Nestor Michael C. Tiglao: "Transport layer caching mechanisms and > optimization", 20+10 min > > Open discussion about the applicability of more precise ECN signalling > > > Cheers, > Michael > > > > > ------------------------------ > > Message: 2 > Date: Sat, 27 Oct 2012 11:26:59 +0200 > From: Michael Welzl <michawe@ifi.uio.no> > Subject: [Iccrg] Why don't we stop treating ECN and loss similarly? > To: iccrg list <iccrg@cs.ucl.ac.uk> > Message-ID: <920D98EE-EFEB-47E2-879C-84999F258771@ifi.uio.no> > Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes > > Hi, > > Here's an idea, inspired by something Bob Briscoe posted to the TSVWG > list recently in a discussion of draft-carlberg-tsvwg-ecn-reactions. > However, this possibly stupid idea is my own responsibility alone :-) > > > According to RFC 3168, senders must react to ECN just as if packets > had been dropped. This is to maintain fairness between ECN-compatible > and non-compatible flows. > Because of this requirement, AQMs cannot ECN-mark packets more > aggressively than it drops packets from non-ECN-capable flows - else > ECN-marked flows would be at a disadvantage. > > We have seen various non-standard congestion control behaviors can co- > exist reasonably well with standard TCP in practice. If it was > possible to have a milder congestion reaction to ECN-based reaction, > it would also be possible to ECN-mark packets earlier, leading to a > bigger advantage for everyone using ECN. And none of this is possible > when we have the "treat an ECN mark just like loss" rule in place. > > Hence, my question: to incentivize ECN usage and enable better > behavior when it's used, shouldn't we remove this rule? > > Note that this is not even about a more fine-grain interpretation of > ECN feedback - it's more like an intermediate step. > > I'm curious what everyone thinks... am I missing something? > > Cheers, > Michael > > > > > ------------------------------ > > _______________________________________________ > Iccrg mailing list > Iccrg@cs.ucl.ac.uk > http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg > > > End of Iccrg Digest, Vol 81, Issue 8 > ************************************ > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://oakham.cs.ucl.ac.uk/pipermail/iccrg/attachments/20121027/1850b1df/attachment.html >From arjuna.sathiaseelan@gmail.com Sat Oct 27 12:16:31 2012 From: arjuna.sathiaseelan@gmail.com (Arjuna Sathiaseelan) Date: Sat Oct 27 12:10:26 2012 Subject: [Iccrg] Re: Iccrg Digest, Vol 81, Issue 8 In-Reply-To: <CAPaG1Akg8qycZkZ+Y_9mKBOAR_K_9MDj0kF_c1gN263-=AHj-A@mail.gmail.com> References: <508bc04d.ef49b40a.25eb.fffffa61SMTPIN_ADDED@mx.google.com> <CAPaG1Akg8qycZkZ+Y_9mKBOAR_K_9MDj0kF_c1gN263-=AHj-A@mail.gmail.com> Message-ID: <CAPaG1AnKSc85O338bqagp8aJ2CbUpzuqohDx3YWhCHXwKsJQeQ@mail.gmail.com> or proportional rate reduction - even better? just random thoughts.. On Oct 27, 2012 12:10 PM, "Arjuna Sathiaseelan" < arjuna.sathiaseelan@gmail.com> wrote: > this sounds great actually..probably reduced to 3/4th cwnd on ecn marks? > > arjuna > On Oct 27, 2012 12:06 PM, <iccrg-request@cs.ucl.ac.uk> wrote: > >> Send Iccrg mailing list submissions to >> iccrg@cs.ucl.ac.uk >> >> To subscribe or unsubscribe via the World Wide Web, visit >> http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg >> or, via email, send a message with subject or body 'help' to >> iccrg-request@cs.ucl.ac.uk >> >> You can reach the person managing the list at >> iccrg-owner@cs.ucl.ac.uk >> >> When replying, please edit your Subject line so it is more specific >> than "Re: Contents of Iccrg digest..." >> >> >> Today's Topics: >> >> 1. Agenda (Michael Welzl) >> 2. Why don't we stop treating ECN and loss similarly? (Michael Welzl) >> >> >> ---------------------------------------------------------------------- >> >> Message: 1 >> Date: Sat, 27 Oct 2012 10:20:13 +0200 >> From: Michael Welzl <michawe@ifi.uio.no> >> Subject: [Iccrg] Agenda >> To: iccrg list <iccrg@cs.ucl.ac.uk> >> Message-ID: <9C0FF8B7-8A75-4599-9980-BBA7B76842F0@ifi.uio.no> >> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes >> >> Hi, >> >> There is still space on our agenda - currently, it looks like this: >> >> >> ICCRG agenda, IETF 85, Atlanta, Monday, 5 November, 13:00-15:00, Salon A >> >> Rong Pan: "A new algorithm for dealing with buffer bloat", 20+10 min >> >> Nestor Michael C. Tiglao: "Transport layer caching mechanisms and >> optimization", 20+10 min >> >> Open discussion about the applicability of more precise ECN signalling >> >> >> Cheers, >> Michael >> >> >> >> >> ------------------------------ >> >> Message: 2 >> Date: Sat, 27 Oct 2012 11:26:59 +0200 >> From: Michael Welzl <michawe@ifi.uio.no> >> Subject: [Iccrg] Why don't we stop treating ECN and loss similarly? >> To: iccrg list <iccrg@cs.ucl.ac.uk> >> Message-ID: <920D98EE-EFEB-47E2-879C-84999F258771@ifi.uio.no> >> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes >> >> Hi, >> >> Here's an idea, inspired by something Bob Briscoe posted to the TSVWG >> list recently in a discussion of draft-carlberg-tsvwg-ecn-reactions. >> However, this possibly stupid idea is my own responsibility alone :-) >> >> >> According to RFC 3168, senders must react to ECN just as if packets >> had been dropped. This is to maintain fairness between ECN-compatible >> and non-compatible flows. >> Because of this requirement, AQMs cannot ECN-mark packets more >> aggressively than it drops packets from non-ECN-capable flows - else >> ECN-marked flows would be at a disadvantage. >> >> We have seen various non-standard congestion control behaviors can co- >> exist reasonably well with standard TCP in practice. If it was >> possible to have a milder congestion reaction to ECN-based reaction, >> it would also be possible to ECN-mark packets earlier, leading to a >> bigger advantage for everyone using ECN. And none of this is possible >> when we have the "treat an ECN mark just like loss" rule in place. >> >> Hence, my question: to incentivize ECN usage and enable better >> behavior when it's used, shouldn't we remove this rule? >> >> Note that this is not even about a more fine-grain interpretation of >> ECN feedback - it's more like an intermediate step. >> >> I'm curious what everyone thinks... am I missing something? >> >> Cheers, >> Michael >> >> >> >> >> ------------------------------ >> >> _______________________________________________ >> Iccrg mailing list >> Iccrg@cs.ucl.ac.uk >> http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg >> >> >> End of Iccrg Digest, Vol 81, Issue 8 >> ************************************ >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://oakham.cs.ucl.ac.uk/pipermail/iccrg/attachments/20121027/5228110e/attachment.html >From john@jlc.net Sat Oct 27 12:55:59 2012 From: john@jlc.net (John Leslie) Date: Sat Oct 27 12:50:13 2012 Subject: [Iccrg] Why don't we stop treating ECN and loss similarly? In-Reply-To: <920D98EE-EFEB-47E2-879C-84999F258771@ifi.uio.no> References: <920D98EE-EFEB-47E2-879C-84999F258771@ifi.uio.no> Message-ID: <20121027115559.GS27557@verdi> Michael Welzl <michawe@ifi.uio.no> wrote: > > According to RFC 3168, senders must react to ECN just as if packets > had been dropped. This is to maintain fairness between ECN-compatible > and non-compatible flows. > Because of this requirement, AQMs cannot ECN-mark packets more > aggressively than it drops packets from non-ECN-capable flows - else > ECN-marked flows would be at a disadvantage. As Bob has pointed out, theory clearly shows that the less-aggressive reaction-to-loss will starve the more-aggressive... > We have seen various non-standard congestion control behaviors can > co-exist reasonably well with standard TCP in practice. Yes! At the risk of setting-off Bob, this suggests that the theory may not completely correspond to reality. ;^) > If it was possible to have a milder congestion reaction to ECN-based > reaction, it would also be possible to ECN-mark packets earlier, > leading to a bigger advantage for everyone using ECN. This, of course, would be trivial if we had a gradation of ECN marking... > And none of this is possible when we have the "treat an ECN mark > just like loss" rule in place. Rules can be relaxed in research projects... > Hence, my question: to incentivize ECN usage and enable better > behavior when it's used, shouldn't we remove this rule? There is certainly room for research which "partially" relaxes this rule. > Note that this is not even about a more fine-grain interpretation of > ECN feedback - it's more like an intermediate step. I'll remind folks that the RMCAT WG needs to convince folks that a less-aggressive response to ECN/loss, coupled with a less- aggressive ramp-up in the absence of ECN/loss, is safe. Early indications are that this will prove acceptable... -- John Leslie <john@jlc.net>
- [Iccrg] Re: Iccrg Digest, Vol 81, Issue 8 Arjuna Sathiaseelan