Message-ID: <AANLkTi=_duV-X26OkDaDB2fcvFJuJSt-hHxpdpG4f=B-@mail.gmail.com>
Date:	Wed, 16 Mar 2011 15:04:57 -0700
From:	Yuchung Cheng <ycheng@...gle.com>
To:	Carsten Wolff <carsten@...ffcarsten.de>
Cc:	David Miller <davem@...emloft.net>,
	Ilpo Jarvinen <ilpo.jarvinen@...sinki.fi>,
	Nandita Dukkipati <nanditad@...gle.com>,
	netdev@...r.kernel.org,
	Alexander Zimmermann <alexander.zimmermann@...sys.rwth-aachen.de>
Subject: Re: [PATCH] tcp: avoid cwnd moderation in undo

Hi Carsten,

Thanks for the detailed explanation. This funnels down to

a) is cwnd moderation always a good idea?

b) if (a) is true, cwnd should be moderated all the time to suppress
any possible burst.

Clearly, Linux code supports (a) and (b). Like John's email, I am not
aware of too much scientific data supporting that.

But I do have a real problem on web server. The HTTP response has
normally a handful of packets. Moderating cwnd at the end or after the
(false) recovery often makes cwnd=3-4. This causes slow-start on the
subsequent HTTP response and throws away some benefit of  persistent
HTTP connections. That slow-start is not necessary and adds some extra
round-trips to serve the tiny object.


On Wed, Mar 16, 2011 at 9:18 AM, Carsten Wolff <carsten@...ffcarsten.de> wrote:
> Hi again,
>
> On Wednesday 16 March 2011, Yuchung Cheng wrote:
>> On Tue, Mar 15, 2011 at 3:07 AM, Carsten Wolff <carsten@...ffcarsten.de>
> wrote:
>> > Hi,
>> >
>> > On Monday 14 March 2011, Yuchung Cheng wrote:
>> > > On Mon, Mar 14, 2011 at 3:06 AM, Carsten Wolff
>> > > <carsten@...ffcarsten.de>
>> >
>> > wrote:
>> > > In the presence of reordering, cwnd is already moderated in Disorder
>> > > state before
>> > >  entering the (false) recovery.
>> >
>> > Sure, cwnd moderation to in_flight + 1 segment is applied in disorder
>> > state,
>>
>> it's in_flight + 3 usually. the moderation first happens
>> tcp_try_to_open() instead of tcp_cwnd_down()
>
> In disorder state, tcp_try_to_open() calls tcp_cwnd_down() which clamps cwnd
> to in_flight + 1 for dupacks (where tcp_packets_in_flight() is not to be
> confused with the IN_FLIGHT variable in IETF documents, which is called
> packets_out in Linux ...). Otherwise, Linux would be violating RFC3042, which
> allows to send one SMSS of data on each dupack before recovery (actually, just
> the first two, but since the DupThresh can be larger than 3 in linux, it
> extends Limited Transmit to more than just the first two dupacks). This is
> mostly equivalent to the aggressive variant of extended limited transmit in
> RFC4653.
>
>> > because this is implementing a form of extended limited transmit.
>> > Nevertheless, after a reordering event that caused a spurious fast
>> > retransmit, there can be an undo of congestion state changes (either
>> > after recovery or interrupting recovery, depending on the options
>> > enabled in the connection). I just wanted to point out, that the
>> > moderation step happening upon an undo may allow a larger burst, if a
>> > previous reordering event was detected and caused tp->reordering to be
>> > increased.
>>
>> Your point is that cwnd should be moderated on reordering (in undo or
>> other events). Point taken.
>>  My point is that cwnd does not need to be moderated on false
>> recoveries. Do you agree?
>> To implement your design, tcp_update_reordering should do
>> tcp_cwnd_moderation().
>> To implement my point, the moderations should be avoided in undo
>> operations.
>>
>> The two aren't in conflict. But there are cases that have both undo
>> and reordering.
>> Are we on the same page?
>
> Unfortunately, no. ;-) My point is, that cwnd should be moderated when the
> congestion state changes are undone after a spurious recovery has been
> detected. Reordering is only one possible reason for a false recovery. And I
> stick to that point because of the thoughts I pointed out in my mail to john,
> i.e. undo typically leading to exceptionally large segment bursts.
>
> As for cwnd moderation upon the detection of a reordering event (that's a
> different thing at a differnt point in time than detection of a false
> recovery!): This wouldn't make sense to me. The detection of the reordering
> event together with a metric that measures the extent of the reordering can be
> used to try and prevent false recoverys in future reordering events, by
> delaying the congestion reaction (i.e. fast retransmit) then.
>
> Reordering can be a cause of spurious recovery. But undo mechanisms and
> mechanisms to prevent false recovery(s) are orthogonal.
>
> Your patch touches all undos, while reordering is just an example for a cause
> of false recovery.
>
>> > > > More importantly, the prior ssthresh is restored and not affected by
>> > > > moderation. This means, if moderation reduces cwnd to a small value,
>> > > > then cwnd < ssthresh and TCP will quickly slow-start back to the
>> > > > previous state, without sending a big burst of segments.
>> >
>> > This is actually the more important point, because it means that the
>> > moderation does not negate the effects of the undo operation, as
>> > suggested by your patch-description.
>>
>> It's a double-edge sword. Why slow-start if there is no real loss?
>
> Its a timing thing. I mean, it is an undo operation: the harm has been done,
> some opportunity to send new data has been lost. Trying to send all that data
> at once now without an ACK-clock will cause more harm when buffers are under
> pressure. The undo operation should not try to make up for lost opportunity,
> only try to reduce further loss of opportunity to send new data. For this, the
> segment bursts have to be moderated.
>
>> It
>> hurts short
>> request-response type traffic performance badly b/c each undo makes cwnd =
>> 3.
>>
>> > False fast retransmits are mostly caused by reordering, spurious RTOs can
>> > also be caused by delay variations that do not exhibit reordering. Your
>> > patch touches all cases of spurious events. Anyway, I just mentioned
>> > reordering, because it is the event in which Linux already allows larger
>> > bursts of size tp->reordering in the moderation function (i.e.
>> > tp->reordering might be increased). It's also not important to me if the
>> > undo is happening duringor after recovery, the important question is, if
>> > burst protection in general is an important goal, or not (and I think
>> > it's there for a reason).
>>
>> I am hoping my previous explanation make sense to you (these two points are
>> not in conflict).
>
> I hope the same for my explanations. :-)
>
> Cheers
> Carsten
> --
>           /\-´-/\
>          (  @ @  )
> ________o0O___^___O0o________
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists