From: Sujith Manoharan <sujith@msujith.org>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Felix Fietkau <nbd@openwrt.org>,
netdev@vger.kernel.org, Dave Taht <dave.taht@gmail.com>
Subject: Re: TCP performance regression
Date: Mon, 11 Nov 2013 21:43:51 +0530 [thread overview]
Message-ID: <21121.575.539384.948990@gargle.gargle.HOWL> (raw)
In-Reply-To: <1384180069.16391.32.camel@edumazet-glaptop2.roam.corp.google.com>
Eric Dumazet wrote:
> We have many choices.
>
> 1) Add back a minimum of ~128 K of outstanding bytes per TCP session,
> so that buggy drivers can sustain 'line rate'.
>
> Note that with 100 concurrent TCP streams, total amount of bytes
> queued on the NIC is 12 MB.
> And pfifo_fast qdisc will drop packets anyway.
>
> Thats what we call 'BufferBloat'
>
> 2) Try lower values like 64K. Still bufferbloat.
>
> 3) Fix buggy drivers, using a proper logic, or shorter timers (mvneta
> case for example)
>
> 4) Add a new netdev attribute, so that well behaving NIC drivers do not
> have to artificially force TCP stack to queue too many bytes in
> Qdisc/NIC queues.
I think the quirks of 802.11 aggregation should be taken into account.
I am adding Felix to this thread, who would have more to say on latency/bufferbloat
with wireless drivers.
Sujith
next prev parent reply other threads:[~2013-11-11 16:18 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-11-11 5:30 TCP performance regression Sujith Manoharan
2013-11-11 5:55 ` Eric Dumazet
2013-11-11 6:07 ` Sujith Manoharan
2013-11-11 6:54 ` Eric Dumazet
2013-11-11 8:19 ` Sujith Manoharan
2013-11-11 14:27 ` Eric Dumazet
2013-11-11 14:39 ` Eric Dumazet
2013-11-11 16:44 ` Eric Dumazet
2013-11-11 15:05 ` David Laight
2013-11-11 15:29 ` Eric Dumazet
2013-11-11 15:43 ` David Laight
2013-11-11 16:17 ` Eric Dumazet
2013-11-11 16:35 ` David Laight
2013-11-11 17:41 ` Eric Dumazet
2013-11-12 7:42 ` Willy Tarreau
2013-11-12 14:16 ` Eric Dumazet
2013-11-14 9:54 ` Dave Taht
2013-11-11 16:13 ` Sujith Manoharan [this message]
2013-11-11 16:38 ` Felix Fietkau
2013-11-11 17:38 ` Eric Dumazet
2013-11-11 17:44 ` Felix Fietkau
2013-11-11 18:03 ` Dave Taht
2013-11-11 18:29 ` Sujith Manoharan
2013-11-11 18:31 ` Dave Taht
2013-11-11 19:11 ` Ben Greear
2013-11-11 19:24 ` Dave Taht
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=21121.575.539384.948990@gargle.gargle.HOWL \
--to=sujith@msujith.org \
--cc=dave.taht@gmail.com \
--cc=eric.dumazet@gmail.com \
--cc=nbd@openwrt.org \
--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;
as well as URLs for NNTP newsgroup(s).