From: Eric Dumazet <eric.dumazet@gmail.com>
To: John Heffner <johnwheffner@gmail.com>
Cc: nanditad@google.com, netdev@vger.kernel.org,
mattmathis@google.com, codel@lists.bufferbloat.net,
ncardwell@google.com, David Miller <davem@davemloft.net>
Subject: Re: [Codel] [RFC PATCH v2] tcp: TCP Small Queues
Date: Thu, 12 Jul 2012 15:46:52 +0200 [thread overview]
Message-ID: <1342100812.3265.8260.camel@edumazet-glaptop> (raw)
In-Reply-To: <CABrhC0=Ls7G-noW1cjsyiF+G5v9f9R=bi6JvrOoT5ZDQB=gSXg@mail.gmail.com>
On Thu, 2012-07-12 at 09:33 -0400, John Heffner wrote:
> One general question: why a per-connection limit? I haven't been
> following the bufferbloat conversation closely so I may have missed
> some of the conversation. But it seems that multiple connections will
> still cause longer queue times.
We already have a per-device limit, in qdisc.
If you want to monitor several tcp sessions, I urge you use a controller
for that. Like codel or fq_codel.
Experiments show that limiting to two TSO packets in qdisc per tcp flow
is enough to stop insane qdisc queueing, without impact on throughput
for people wanting fast tcp sessions.
Thats not solving the more general problem of having 1000 competing
flows.
next prev parent reply other threads:[~2012-07-12 13:46 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-06-28 17:07 [Codel] [PATCH net-next] fq_codel: report congestion notification at enqueue time Eric Dumazet
2012-06-28 17:51 ` Dave Taht
2012-06-28 18:12 ` Eric Dumazet
2012-06-28 22:56 ` Yuchung Cheng
2012-06-28 23:47 ` Dave Taht
2012-06-29 4:50 ` Eric Dumazet
2012-06-29 5:24 ` Dave Taht
2012-07-04 10:11 ` [Codel] [RFC PATCH] tcp: limit data skbs in qdisc layer Eric Dumazet
2012-07-09 7:08 ` David Miller
2012-07-09 8:03 ` Eric Dumazet
2012-07-09 8:48 ` Eric Dumazet
2012-07-09 14:55 ` Eric Dumazet
2012-07-10 13:28 ` Lin Ming
2012-07-10 15:13 ` [Codel] [RFC PATCH v2] tcp: TCP Small Queues Eric Dumazet
2012-07-10 17:06 ` Eric Dumazet
2012-07-10 17:37 ` Yuchung Cheng
2012-07-10 18:32 ` Eric Dumazet
2012-07-11 15:11 ` Eric Dumazet
2012-07-11 15:16 ` Ben Greear
2012-07-11 15:25 ` Eric Dumazet
2012-07-11 15:43 ` Ben Greear
2012-07-11 15:54 ` Eric Dumazet
2012-07-11 16:03 ` Ben Greear
2012-07-11 18:23 ` Rick Jones
2012-07-11 23:38 ` Eric Dumazet
2012-07-11 18:44 ` Rick Jones
2012-07-11 23:49 ` Eric Dumazet
2012-07-12 7:34 ` Eric Dumazet
2012-07-12 7:37 ` David Miller
2012-07-12 7:51 ` Eric Dumazet
2012-07-12 14:55 ` Tom Herbert
2012-07-12 13:33 ` John Heffner
2012-07-12 13:46 ` Eric Dumazet [this message]
2012-07-12 16:44 ` John Heffner
2012-07-12 16:54 ` Jim Gettys
2012-06-28 23:52 ` [Codel] [PATCH net-next] fq_codel: report congestion notification at enqueue time Nandita Dukkipati
2012-06-29 4:18 ` Eric Dumazet
2012-06-29 4:53 ` Eric Dumazet
2012-06-29 5:12 ` David Miller
2012-06-29 5:24 ` Eric Dumazet
2012-06-29 5:29 ` David Miller
2012-06-29 5:50 ` Eric Dumazet
2012-06-29 7:53 ` David Miller
2012-06-29 8:04 ` David Miller
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
List information: https://lists.bufferbloat.net/postorius/lists/codel.lists.bufferbloat.net/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1342100812.3265.8260.camel@edumazet-glaptop \
--to=eric.dumazet@gmail.com \
--cc=codel@lists.bufferbloat.net \
--cc=davem@davemloft.net \
--cc=johnwheffner@gmail.com \
--cc=mattmathis@google.com \
--cc=nanditad@google.com \
--cc=ncardwell@google.com \
--cc=netdev@vger.kernel.org \
/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