Date: Fri, 18 Jan 2019 02:15:12 -0800
From: Brian Trammell <notifications@github.com>
Reply-To: quicwg/base-drafts
 <reply+0166e4abd9c762b5303f751eb634c984f5686af5d8a86e9492cf000000011859693092a169ce17e0dd5b@reply.github.com>
To: quicwg/base-drafts <base-drafts@noreply.github.com>
Cc: Subscribed <subscribed@noreply.github.com>
Message-ID: <quicwg/base-drafts/issues/2348/455496194@github.com>
In-Reply-To: <quicwg/base-drafts/issues/2348@github.com>
References: <quicwg/base-drafts/issues/2348@github.com>
Subject: Re: [quicwg/base-drafts] Specify IPv6 flow label for QUIC (#2348)
Mime-Version: 1.0
Content-Type: multipart/alternative;
 boundary="--==_mimepart_5c41a73026d9a_cf43fabdaad45c411294d";
 charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: list
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic-issues/MA11_SzfCcG-V2CEtvfs_n8LkSE>


----==_mimepart_5c41a73026d9a_cf43fabdaad45c411294d
Content-Type: text/plain;
 charset=UTF-8
Content-Transfer-Encoding: 7bit

Probably. There are (probably minor) caveats here.

If/when [RFC 6437](https://tools.ietf.org/html/6437) is widely deployed in the forwarding plane (I have neither data nor intuition about this), this will cause ECMP to operate on the DCID, so switching DCID will cause packets to be grouped along a different path. 

The upside of this is that in cases where ECMP is balanced across different _interdomain_ paths (this does happen, though I'm not sure how rarely -- though I know of recent work on traceroute enhancements to compensate for increasingly complex multipath topologies in the Internet), the cost of timing-correlation linkability across DCID change goes way up in the Internet core.

The downside is that a DCID switch will be more likely to cause latency discontinuity; when the switch hashes on to a faster path, there will be a spike in reordering at the receiver. This won't break anything per se (QUIC handles reordering just fine as specified), but it may lead to adverse performance in the few windows around a DCID switch that isn't currently there.

tl;dr it's certainly worth doing, but let's not be surprised by the (minor) effects of the decision.

-- 
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/2348#issuecomment-455496194
----==_mimepart_5c41a73026d9a_cf43fabdaad45c411294d
Content-Type: text/html;
 charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<p>Probably. There are (probably minor) caveats here.</p>
<p>If/when <a href=3D"https://tools.ietf.org/html/6437" rel=3D"nofollow">=
RFC 6437</a> is widely deployed in the forwarding plane (I have neither d=
ata nor intuition about this), this will cause ECMP to operate on the DCI=
D, so switching DCID will cause packets to be grouped along a different p=
ath.</p>
<p>The upside of this is that in cases where ECMP is balanced across diff=
erent <em>interdomain</em> paths (this does happen, though I'm not sure h=
ow rarely -- though I know of recent work on traceroute enhancements to c=
ompensate for increasingly complex multipath topologies in the Internet),=
 the cost of timing-correlation linkability across DCID change goes way u=
p in the Internet core.</p>
<p>The downside is that a DCID switch will be more likely to cause latenc=
y discontinuity; when the switch hashes on to a faster path, there will b=
e a spike in reordering at the receiver. This won't break anything per se=
 (QUIC handles reordering just fine as specified), but it may lead to adv=
erse performance in the few windows around a DCID switch that isn't curre=
ntly there.</p>
<p>tl;dr it's certainly worth doing, but let's not be surprised by the (m=
inor) effects of the decision.</p>

<p style=3D"font-size:small;-webkit-text-size-adjust:none;color:#666;">&m=
dash;<br />You are receiving this because you are subscribed to this thre=
ad.<br />Reply to this email directly, <a href=3D"https://github.com/quic=
wg/base-drafts/issues/2348#issuecomment-455496194">view it on GitHub</a>,=
 or <a href=3D"https://github.com/notifications/unsubscribe-auth/AWbkqz_h=
szoy1gVIatQjWnHedWIOYMrYks5vEZ6wgaJpZM4aHaIS">mute the thread</a>.<img sr=
c=3D"https://github.com/notifications/beacon/AWbkqyyBKzMYt4XPzZRRdQ4u3jC2=
w7FCks5vEZ6wgaJpZM4aHaIS.gif" height=3D"1" width=3D"1" alt=3D"" /></p>
<script type=3D"application/json" data-scope=3D"inboxmarkup">{"api_versio=
n":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name"=
:"GitHub"},"entity":{"external_key":"github/quicwg/base-drafts","title":"=
quicwg/base-drafts","subtitle":"GitHub repository","main_image_url":"http=
s://github.githubassets.com/images/email/message_cards/header.png","avata=
r_image_url":"https://github.githubassets.com/images/email/message_cards/=
avatar.png","action":{"name":"Open in GitHub","url":"https://github.com/q=
uicwg/base-drafts"}},"updates":{"snippets":[{"icon":"PERSON","message":"@=
britram in #2348: Probably. There are (probably minor) caveats here.\r\n\=
r\nIf/when [RFC 6437](https://tools.ietf.org/html/6437) is widely deploye=
d in the forwarding plane (I have neither data nor intuition about this),=
 this will cause ECMP to operate on the DCID, so switching DCID will caus=
e packets to be grouped along a different path. \r\n\r\nThe upside of thi=
s is that in cases where ECMP is balanced across different _interdomain_ =
paths (this does happen, though I'm not sure how rarely -- though I know =
of recent work on traceroute enhancements to compensate for increasingly =
complex multipath topologies in the Internet), the cost of timing-correla=
tion linkability across DCID change goes way up in the Internet core.\r\n=
\r\nThe downside is that a DCID switch will be more likely to cause laten=
cy discontinuity; when the switch hashes on to a faster path, there will =
be a spike in reordering at the receiver. This won't break anything per s=
e (QUIC handles reordering just fine as specified), but it may lead to ad=
verse performance in the few windows around a DCID switch that isn't curr=
ently there.\r\n\r\ntl;dr it's certainly worth doing, but let's not be su=
rprised by the (minor) effects of the decision."}],"action":{"name":"View=
 Issue","url":"https://github.com/quicwg/base-drafts/issues/2348#issuecom=
ment-455496194"}}}</script>
<script type=3D"application/ld+json">[
{
"@context": "http://schema.org",
"@type": "EmailMessage",
"potentialAction": {
"@type": "ViewAction",
"target": "https://github.com/quicwg/base-drafts/issues/2348#issuecomment=
-455496194",
"url": "https://github.com/quicwg/base-drafts/issues/2348#issuecomment-45=
5496194",
"name": "View Issue"
},
"description": "View this Issue on GitHub",
"publisher": {
"@type": "Organization",
"name": "GitHub",
"url": "https://github.com"
}
}
]</script>=

----==_mimepart_5c41a73026d9a_cf43fabdaad45c411294d--