IETF Logo Mail Archive

Re: [quicwg/base-drafts] QUIC PTO is too conservative, causing a measurable regression in tail latency (#3526)

ianswett <notifications@github.com> Tue, 12 May 2020 14:48 UTC

Return-Path: <noreply@github.com>
X-Original-To: quic-issues@ietfa.amsl.com
Delivered-To: quic-issues@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DB5B53A0ABE for <quic-issues@ietfa.amsl.com>; Tue, 12 May 2020 07:48:08 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.634
X-Spam-Level:
X-Spam-Status: No, score=-1.634 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.173, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_IMAGE_ONLY_20=1.546, HTML_MESSAGE=0.001, HTML_OBFUSCATE_10_20=0.093, MAILING_LIST_MULTI=-1, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=no autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=github.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id UyzXqc2cfdr5 for <quic-issues@ietfa.amsl.com>; Tue, 12 May 2020 07:48:07 -0700 (PDT)
Received: from out-20.smtp.github.com (out-20.smtp.github.com [192.30.252.203]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id A88873A0A99 for <quic-issues@ietf.org>; Tue, 12 May 2020 07:48:07 -0700 (PDT)
Received: from github-lowworker-9bcb4a1.ac4-iad.github.net (github-lowworker-9bcb4a1.ac4-iad.github.net [10.52.25.84]) by smtp.github.com (Postfix) with ESMTP id 969B28C05A5 for <quic-issues@ietf.org>; Tue, 12 May 2020 07:48:06 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=github.com; s=pf2014; t=1589294886; bh=bmqmv+Rr1Pyeek3xDSDbKBXSTsdSMseusOuBP9d9U2o=; h=Date:From:Reply-To:To:Cc:In-Reply-To:References:Subject:List-ID: List-Archive:List-Post:List-Unsubscribe:From; b=bQOjm2B4lauLleI53zCMfkiH7UdJ12mqTHu3Z2PQAo+1+fiRzIrehc/W257fwKOqG qRvf+qmdJXm5W+KkhhNzgtgeVkfUy26DA2Ts/NZCAHT/iZaircRCmXpwjru6vx3XeY F7ZGyhqWLchkPk9RXZJwk+PvsmAuogrQE5ZvqAAo=
Date: Tue, 12 May 2020 07:48:06 -0700
From: ianswett <notifications@github.com>
Reply-To: quicwg/base-drafts <reply+AFTOJK6MMBTK6IY5DBZPL554Y2MCNEVBNHHCFNKKLA@reply.github.com>
To: quicwg/base-drafts <base-drafts@noreply.github.com>
Cc: Subscribed <subscribed@noreply.github.com>
Message-ID: <quicwg/base-drafts/issues/3526/627392539@github.com>
In-Reply-To: <quicwg/base-drafts/issues/3526@github.com>
References: <quicwg/base-drafts/issues/3526@github.com>
Subject: Re: [quicwg/base-drafts] QUIC PTO is too conservative, causing a measurable regression in tail latency (#3526)
Mime-Version: 1.0
Content-Type: multipart/alternative; boundary="--==_mimepart_5ebab72687859_66403f82930cd960106321"; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Precedence: list
X-GitHub-Sender: ianswett
X-GitHub-Recipient: quic-issues
X-GitHub-Reason: subscribed
X-Auto-Response-Suppress: All
X-GitHub-Recipient-Address: quic-issues@ietf.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic-issues/6dHtgFRcplvJAkwV53eirrNKwK8>
X-BeenThere: quic-issues@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Notification list for GitHub issues related to the QUIC WG <quic-issues.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic-issues>, <mailto:quic-issues-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic-issues/>
List-Post: <mailto:quic-issues@ietf.org>
List-Help: <mailto:quic-issues-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic-issues>, <mailto:quic-issues-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 12 May 2020 14:48:09 -0000

The most recent set of experiments indicate that my above suggestion of arming for max(left_edge + SRTT+4*RTTVar, right_edge + 1.5*SRTT) is a very slight improvement for some metrics, but it looks to be in the noise for most metrics.

We did find SRTT+2*RTTVar is as good or better than SRTT+4*RTTVar for all application metrics, with a very small difference in retransmit rates(probably because PTO can send new data).

I'd be very curious to see what @mjoras finds in his testing, but I'm increasingly inclined to close this with no action.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/3526#issuecomment-627392539

v2.23.0 | Report a Bug | By Email