public inbox for netdev@vger.kernel.org 
 help / color / mirror / Atom feed
From: Corey Hickey <bugfood-ml@fatooh•org>
To: Christoph Paasch <christoph.paasch@uclouvain•be>
Cc: Eric Dumazet <eric.dumazet@gmail•com>,
	netdev@vger•kernel.org, Phil Oester <kernel@linuxace•com>,
	Benjamin Hesmans <benjamin.hesmans@uclouvain•be>
Subject: Re: [RFC 0/2] Account for duplicate ACKs with invalid SACK-blocks
Date: Mon, 19 Aug 2013 14:22:46 -0700	[thread overview]
Message-ID: <52128CA6.6090001@fatooh.org> (raw)
In-Reply-To: <1376940568-16512-1-git-send-email-christoph.paasch@uclouvain.be>

On 2013-08-19 12:29, Christoph Paasch wrote:
> There exist sequence-number rewriting middleboxes, who do not modify the 
> sequence-number in the SACK-blocks.
> Duplicate acknowledgments with these (invalid) SACK-blocks will not be
> accounted in sacked_out, and thus no fast-retransmit will trigger.
> So, the only way to recover from a packet-loss is through an RTO,
> effectively killing the performance of TCP in this case.
> 
> Performance-results can be seen here:
> http://tools.ietf.org/agenda/87/slides/slides-87-tcpm-11.pdf
> 
> Another solution might be to simply disable SACK, as soon as an invalid
> SACK-block has been received (as suggested by Phil Oester). This however
> might be too aggressive.
> 
> Christoph Paasch (2):
>   Use acked_out for reno-style ack acounting instead of sacked_out
>   Account acked_out in sack, if the sack is invalid
> 
>  include/linux/tcp.h      |   1 +
>  include/net/tcp.h        |  17 +++++---
>  net/ipv4/tcp_input.c     | 103 ++++++++++++++++++++++++++++++++---------------
>  net/ipv4/tcp_minisocks.c |   1 +
>  net/ipv4/tcp_output.c    |   6 +--
>  net/ipv4/tcp_timer.c     |   2 +-
>  6 files changed, 89 insertions(+), 41 deletions(-)

I tested this patchset, and, unfortunately, the behavior seems to be the
same.

I'm still planning on working with the Cisco device, when I can get some
of the network admin's time, but that won't happen immediately.

Thanks,
Corey

  parent reply	other threads:[~2013-08-19 21:22 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-08-18  5:55 NAT stops forwarding ACKs after PMTU discovery Corey Hickey
2013-08-18 15:24 ` Eric Dumazet
2013-08-18 16:59   ` Corey Hickey
2013-08-18 21:23   ` Jozsef Kadlecsik
2013-08-19  0:00     ` Eric Dumazet
2013-08-19  0:03       ` Eric Dumazet
2013-08-19  8:43         ` Corey Hickey
2013-08-19 12:33           ` Christoph Paasch
2013-08-19 13:24             ` Eric Dumazet
2013-08-19 13:49               ` Christoph Paasch
2013-08-19 13:58                 ` Eric Dumazet
2013-08-19 14:35                   ` Phil Oester
2013-08-19 15:32                     ` Eric Dumazet
2013-08-19 15:33                     ` Christoph Paasch
2013-08-19 16:00                       ` Eric Dumazet
2013-08-19 17:15                         ` Christoph Paasch
2013-08-19 18:00                           ` Phil Oester
2013-08-19 18:10                           ` Eric Dumazet
2013-08-19 19:29                             ` [RFC 0/2] Account for duplicate ACKs with invalid SACK-blocks Christoph Paasch
2013-08-19 19:29                               ` [RFC 1/2] Use acked_out for reno-style ack acounting instead of sacked_out Christoph Paasch
2013-08-19 19:29                               ` [RFC 2/2] Account acked_out in sack, if the sack is invalid Christoph Paasch
2013-08-19 21:22                               ` Corey Hickey [this message]
2013-08-20  7:36                                 ` [RFC 0/2] Account for duplicate ACKs with invalid SACK-blocks Christoph Paasch
2013-08-22  3:32                               ` David Miller
2013-08-22  4:15                                 ` Corey Hickey
2013-08-19 14:43                   ` NAT stops forwarding ACKs after PMTU discovery Christoph Paasch
2013-08-19 20:13                   ` Jozsef Kadlecsik
2013-08-19 20:43                     ` Christoph Paasch
2013-08-19 21:08                     ` Eric Dumazet
2013-08-19 22:07                       ` Jozsef Kadlecsik
2013-08-20  4:18                         ` Corey Hickey
2013-08-19 18:22             ` Corey Hickey

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=52128CA6.6090001@fatooh.org \
    --to=bugfood-ml@fatooh$(echo .)org \
    --cc=benjamin.hesmans@uclouvain$(echo .)be \
    --cc=christoph.paasch@uclouvain$(echo .)be \
    --cc=eric.dumazet@gmail$(echo .)com \
    --cc=kernel@linuxace$(echo .)com \
    --cc=netdev@vger$(echo .)kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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