From: Eric Dumazet <eric.dumazet@gmail.com>
To: Jana Iyengar <jri.ietf@gmail.com>, Ian Swett <ianswett@google.com>
Cc: Christoph Paasch <cpaasch@apple.com>,
willemdebruijn.kernel@gmail.com, netdev@vger.kernel.org,
lhedstrom@apple.com
Subject: Re: [RFC 0/2] Delayed binding of UDP sockets for Quic per-connection sockets
Date: Sun, 4 Nov 2018 10:58:43 -0800 [thread overview]
Message-ID: <fb15e4df-83c6-2c32-c3c3-392d2e66a378@gmail.com> (raw)
In-Reply-To: <CACpbDccs6WmLCknpu2GLMMBnkHwS4apsr3Z3sAKt4Ch_2HPwgg@mail.gmail.com>
On 11/04/2018 02:45 AM, Jana Iyengar wrote:
> I think SO_TXTIME solves the most egregious problem I have with using sched_fq for QUIC. That's a great help, so thank you!
>
> That said, as Ian says, SO_TXTIME does not allow for the flow isolation properties of sched_fq, which would be a nice secondary benefit. I suspect that can also be done in a similar manner to SO_TXTIME -- by attaching an opaque label to each sendmsg which is used by sched_fq to determine the flow. Is that feasible?
The plan is to have flow isolation, without having to change applications to provide a flow identifier,
since we can not trust user applications anyway.
FQ will perform a proper flow dissection for packets sent over unconnected UDP sockets.
FQ has two parts [1], one being used by locally generated packets (local TCP stack),
one being used in forwarding workloads.
The patch which I will send shortly (when net-next reopens)
will only make sure FQ does not use skb->sk as a flow identifier
if the socket is not a connected one.
[1] See commit 06eb395fa9856b5a87cf7d80baee2a0ed3cdb9d7 for some details.
prev parent reply other threads:[~2018-11-05 4:14 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-31 23:26 [RFC 0/2] Delayed binding of UDP sockets for Quic per-connection sockets Christoph Paasch
2018-10-31 23:26 ` [RFC 1/2] net: Add new socket-option SO_DELAYED_BIND Christoph Paasch
2018-10-31 23:26 ` [RFC 2/2] udp: Support SO_DELAYED_BIND Christoph Paasch
2018-11-01 0:53 ` [RFC 0/2] Delayed binding of UDP sockets for Quic per-connection sockets Eric Dumazet
2018-11-01 3:50 ` Christoph Paasch
2018-11-01 5:04 ` Eric Dumazet
2018-11-01 5:07 ` Christoph Paasch
2018-11-01 5:08 ` Eric Dumazet
2018-11-01 5:17 ` Eric Dumazet
2018-11-01 17:58 ` Leif Hedstrom
2018-11-01 18:21 ` Eric Dumazet
2018-11-01 21:51 ` Willem de Bruijn
2018-11-01 22:11 ` Christoph Paasch
[not found] ` <CAKcm_gNZqgRGRj2J5yJDsavHsoaeXtozrbGp+TmAj_DRsCUOLQ@mail.gmail.com>
[not found] ` <CACpbDccs6WmLCknpu2GLMMBnkHwS4apsr3Z3sAKt4Ch_2HPwgg@mail.gmail.com>
2018-11-04 18:58 ` Eric Dumazet [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=fb15e4df-83c6-2c32-c3c3-392d2e66a378@gmail.com \
--to=eric.dumazet@gmail.com \
--cc=cpaasch@apple.com \
--cc=ianswett@google.com \
--cc=jri.ietf@gmail.com \
--cc=lhedstrom@apple.com \
--cc=netdev@vger.kernel.org \
--cc=willemdebruijn.kernel@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).