Re: [quicwg/base-drafts] MSS Clamping Support (#3330)

Matt Corallo <> Thu, 09 January 2020 06:34 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 14B3A120137 for <>; Wed, 8 Jan 2020 22:34:51 -0800 (PST)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -8
X-Spam-Status: No, score=-8 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=0.001, MAILING_LIST_MULTI=-1, RCVD_IN_DNSWL_HI=-5, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: (amavisd-new); dkim=pass (1024-bit key)
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id 48aZ8VQo_UPP for <>; Wed, 8 Jan 2020 22:34:49 -0800 (PST)
Received: from ( []) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by (Postfix) with ESMTPS id 70678120118 for <>; Wed, 8 Jan 2020 22:34:49 -0800 (PST)
Date: Wed, 08 Jan 2020 22:34:48 -0800
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=pf2014; t=1578551688; bh=yIF9z6f3336/wXQB4kjeGQTXFGyF1/mGEqlSbIT2rko=; h=Date:From:Reply-To:To:Cc:In-Reply-To:References:Subject:List-ID: List-Archive:List-Post:List-Unsubscribe:From; b=L5NmU9bUUDekdEBl6PvYVKn52P0b0DkMxrdpqmt5SIPyux3P6wsm53JRfqSNLu44t 3pvs41fAM69/MfgWZMzkKGTwgU9t8FqrXG8sK0IpbTf4pXRxYj8ONYPdMirqdsPidG APCEvXJAlQTf6gBiQNl9mnKTA71jIKzds23SLyOE=
From: Matt Corallo <>
Reply-To: quicwg/base-drafts <>
To: quicwg/base-drafts <>
Cc: Subscribed <>
Message-ID: <quicwg/base-drafts/issues/3330/>
In-Reply-To: <quicwg/base-drafts/issues/>
References: <quicwg/base-drafts/issues/>
Subject: Re: [quicwg/base-drafts] MSS Clamping Support (#3330)
Mime-Version: 1.0
Content-Type: multipart/alternative; boundary="--==_mimepart_5e16c988887bb_6d233ff610acd96010392c"; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Precedence: list
X-GitHub-Sender: TheBlueMatt
X-GitHub-Recipient: quic-issues
X-GitHub-Reason: subscribed
X-Auto-Response-Suppress: All
Archived-At: <>
X-Mailman-Version: 2.1.29
List-Id: Notification list for GitHub issues related to the QUIC WG <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 09 Jan 2020 06:34:51 -0000

Right, the question is in large part motivated by the common wisdom that “PMTUD just doesn’t work well enough to be relied on “, which appears to match my own experience - too many poorly-designed middleboxes drop ICMP (cause, you know, security, or something). I suppose this means in practice things will just run at 1280 with timeout-based PMTUD for long-running flows/common endpoints?

> On Jan 9, 2020, at 01:29, Martin Thomson <> wrote:
> The section on PMTUD addresses this point. In addition to PLPMTUD, we also describe how to use ICMP.
> The answer we have does allow the path to send signals, but it doesn't allow for packet modification in any way. In short, if the ICMP message gets through AND it contains enough information that the endpoint can be assured that the sender was on-path, then it can be used.
> That's probably not an especially satisfactory answer, but I believe that is the state of the protocol and at this stage it would require significant effort to change - in the core protocol.
> If there is a high path MTU but smaller packets would exhibit better performance, we currently have no answer. That doesn't mean it is impossible to design a solution, but that would have to go through the endpoints and we don't currently have a way to do that. One would need to be defined. I suspect that the chances of getting that in for QUIC version 1 approaches zero, but this is work that can proceed independently.
> Defining a generic IP-layer capability that protocols like QUIC could benefit from would be my suggested approach. After all, I'm sure that the WebRTC folks would be thrilled to learn how to reduce latency or throughput for video on links that have these characteristics.
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub, or unsubscribe.

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub: