Re: [tcpm] finalizing CUBIC draft (chairs' view)

Jonathan Morton <> Thu, 13 October 2022 11:54 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 93D7DC1524B5; Thu, 13 Oct 2022 04:54:08 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -6.858
X-Spam-Status: No, score=-6.858 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_HI=-5, RCVD_IN_ZEN_BLOCKED_OPENDNS=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01] autolearn=ham autolearn_force=no
Authentication-Results: (amavisd-new); dkim=pass (2048-bit key)
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id hfqTkHoLRpwL; Thu, 13 Oct 2022 04:54:06 -0700 (PDT)
Received: from ( [IPv6:2a00:1450:4864:20::235]) (using TLSv1.3 with cipher TLS_AES_128_GCM_SHA256 (128/128 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by (Postfix) with ESMTPS id 1C82EC14F6EC; Thu, 13 Oct 2022 04:54:06 -0700 (PDT)
Received: by with SMTP id bs14so1992364ljb.9; Thu, 13 Oct 2022 04:54:06 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=20210112; h=to:references:message-id:content-transfer-encoding:cc:date :in-reply-to:from:subject:mime-version:from:to:cc:subject:date :message-id:reply-to; bh=AbVPmh1+CjCKGeeXPz1Ue3K/RyAL3mHIbjH9t2KoRTg=; b=jn/Zhvifw3TRlrb4TEml9fXj9oYoVKI3N/6ZuzZEf9XLnXbwFM1R3G2ewsH9SXsfnO LbGXCGGvs1AXQVBiVVdYYnrzcXEnAJeA5DH4+5a7EJ3WL6L87/Tc2LbecN4bWalUDJnK k8jXs1mlutsW6QZUdXvInu9U/7lcldJNKq6olrEcrnB0fMzmO1CxutgAiNPJkjTBR+u7 VRy0FTCGO18fGo23zYsP8Q3DGCC7WQKu3abfB4s1v15LVz/fiKqcLh5KqaZ+gU1avSDW dNpQlpfB1XuhTQL4TzYHXOXQyBwWsq3iHEHospSPcDnZ8oM8gcKCjKz8BZiQmmfiu3KH caNw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=20210112; h=to:references:message-id:content-transfer-encoding:cc:date :in-reply-to:from:subject:mime-version:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=AbVPmh1+CjCKGeeXPz1Ue3K/RyAL3mHIbjH9t2KoRTg=; b=2BFWzT+9ik5ABU24ytCoSP1GtHUK0bWO91k4kus2M0D5SyZBrK+o6m/R/JlAcSjJ0E mGKjjKrh17KMRICJz61O1mCizg/n3RzBF2MGYboGE2+oCSb1LN+KPXHwNYc46FL6bhBE xExuvVmKDnZ+X1kXWSr7HqdWnacWmxOdHQHn/FiJGJh+OZf2pJp7t6mWuQeuAlchGp38 K7KheMspnL4U7CgqBbQbx+JjysSlVTxk4ALku3jwA8uuJv88aydC4pFbzX1diJtEqQTm LD0WP4Vhwlmqy/aQjl706DptkHXbaB2TYLvaOUVtAbClqeuKqIqjtEKQcq1kVRQPfA7o TTWw==
X-Gm-Message-State: ACrzQf1MzAXqr5WmGtTv85q8vyiEO7ofR0+KZAAQeBqjiOEkkpsYgtV5 9TmrwG/7ASE3BsgG9PKoZlQ=
X-Google-Smtp-Source: AMsMyM6Kb7o2yZo4oRztRpwipAdcwUbhFITfE818pg4SfdO5drxuNF6L1t+4Uq8RSlIH3srpgiGQiQ==
X-Received: by 2002:a05:651c:2212:b0:26c:2baf:652e with SMTP id y18-20020a05651c221200b0026c2baf652emr13569504ljq.84.1665662043702; Thu, 13 Oct 2022 04:54:03 -0700 (PDT)
Received: from ( []) by with ESMTPSA id bg26-20020a05651c0b9a00b0026bfc8d4bbbsm427401ljb.125.2022. (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Thu, 13 Oct 2022 04:54:03 -0700 (PDT)
Content-Type: text/plain; charset="us-ascii"
Mime-Version: 1.0 (Mac OS X Mail 14.0 \(3654.\))
From: Jonathan Morton <>
In-Reply-To: <>
Date: Thu, 13 Oct 2022 14:54:01 +0300
Cc: Markku Kojo <>, Yoshifumi Nishida <>, " Extensions" <>, tcpm-chairs <>
Content-Transfer-Encoding: quoted-printable
Message-Id: <>
References: <> <> <> <> <> <>
To: Ian Swett <>
X-Mailer: Apple Mail (2.3654.
Archived-At: <>
Subject: Re: [tcpm] finalizing CUBIC draft (chairs' view)
X-Mailman-Version: 2.1.39
Precedence: list
List-Id: TCP Maintenance and Minor Extensions Working Group <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 13 Oct 2022 11:54:08 -0000

> On 13 Oct, 2022, at 1:47 pm, Ian Swett <> wrote:
> IMO the question should be whether deploying Cubic has or will cause measurable harm on the public internet vs Reno.  I have never seen evidence of this because I think it's so rare it's in practice unmeasurable.

You're probably correct that the effect is not significant in practice.  However, there is a debate on that point and data has been requested to support that debate, so it is fortuitous that we are in a position to supply that data in the near future.

> Unfairness is only a problem if it's causing latency or drops to other flows (aka externalities).  If you're running a very unfair congestion controller on a dedicated link and it's the only flow, there's no harm.  If the flow is constantly app-limited or never exits slow start (the vast majority of flows), convergence in congestion avoidance is irrelevant.  How it exits slow start actually ends up mattering more, and Cubic recommends Hystart++.

A confounding factor in this analysis is that the potential bad behaviour is generally associated with dynamic conditions, which occur only transiently, rather than steady state conditions which may occur for long periods of observation.  Transient conditions are notoriously difficult to measure and reason about "in the wild", but much easier to induce and measure in the lab, from which we can then extrapolate to likely real-world conditions.

But you are perfectly correct that it is the externalities that matter most.  That is why I mentioned using the "harm metric" to help quantify them.  Under lab conditions it is straightforward to collect the appropriate comparative data from which harm metrics can be computed accurately.

 - Jonathan Morton