Date: Wed, 29 Jul 2020 09:48:35 -0700
From: Jana Iyengar <notifications@github.com>
Reply-To: quicwg/base-drafts
 <reply+AFTOJK5TQYROOF2GFMRD3HF5FWEWHEVBNHHCPUC7GA@reply.github.com>
To: quicwg/base-drafts <base-drafts@noreply.github.com>
Cc: Subscribed <subscribed@noreply.github.com>
Message-ID: <quicwg/base-drafts/issues/3972/665777673@github.com>
In-Reply-To: <quicwg/base-drafts/issues/3972@github.com>
References: <quicwg/base-drafts/issues/3972@github.com>
Subject: Re: [quicwg/base-drafts] Persistent congestion period: text and
 pseudo-code don't agree (#3972)
Mime-Version: 1.0
Content-Type: multipart/alternative;
 boundary="--==_mimepart_5f21a86322bc6_1a6c16f830048e";
 charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: list
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic-issues/c9Xr5Ync5m7TU3TkDhcP-VjpkIc>


----==_mimepart_5f21a86322bc6_1a6c16f830048e
Content-Type: text/plain;
 charset=UTF-8
Content-Transfer-Encoding: 7bit

Option 1 and 2 have a behavior difference with the following implication:

With option 2, a sender might mark persistent congestion with fewer PTOs than if the sender had a more accurate RTT estimate. This is arguably not a significant concern, since:
- the sender will still have sent at least two packets in that duration and marked them lost, since you need two ack-eliciting packets to be lost to declare persistent congestion; and
- the number of packets sent is also sensitive to how the application sends, and a sender cannot rely on that anyway.

Option 2 has the benefit that it is simpler to implement.

On balance, I think Option 2 is better, though it does make this a design change.

-- 
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/3972#issuecomment-665777673
----==_mimepart_5f21a86322bc6_1a6c16f830048e
Content-Type: text/html;
 charset=UTF-8
Content-Transfer-Encoding: 7bit

<p></p>
<p>Option 1 and 2 have a behavior difference with the following implication:</p>
<p>With option 2, a sender might mark persistent congestion with fewer PTOs than if the sender had a more accurate RTT estimate. This is arguably not a significant concern, since:</p>
<ul>
<li>the sender will still have sent at least two packets in that duration and marked them lost, since you need two ack-eliciting packets to be lost to declare persistent congestion; and</li>
<li>the number of packets sent is also sensitive to how the application sends, and a sender cannot rely on that anyway.</li>
</ul>
<p>Option 2 has the benefit that it is simpler to implement.</p>
<p>On balance, I think Option 2 is better, though it does make this a design change.</p>

<p style="font-size:small;-webkit-text-size-adjust:none;color:#666;">&mdash;<br />You are receiving this because you are subscribed to this thread.<br />Reply to this email directly, <a href="https://github.com/quicwg/base-drafts/issues/3972#issuecomment-665777673">view it on GitHub</a>, or <a href="https://github.com/notifications/unsubscribe-auth/AFTOJK4NFUBTBSYJ45GS55DR6BHGHANCNFSM4PLYL7WA">unsubscribe</a>.<img src="https://github.com/notifications/beacon/AFTOJK2OMFLQBUBTIUIW4M3R6BHGHA5CNFSM4PLYL7WKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOE6XPMCI.gif" height="1" width="1" alt="" /></p>
<script type="application/ld+json">[
{
"@context": "http://schema.org",
"@type": "EmailMessage",
"potentialAction": {
"@type": "ViewAction",
"target": "https://github.com/quicwg/base-drafts/issues/3972#issuecomment-665777673",
"url": "https://github.com/quicwg/base-drafts/issues/3972#issuecomment-665777673",
"name": "View Issue"
},
"description": "View this Issue on GitHub",
"publisher": {
"@type": "Organization",
"name": "GitHub",
"url": "https://github.com"
}
}
]</script>
----==_mimepart_5f21a86322bc6_1a6c16f830048e--