Cake - FQ_codel the next generation
 help / color / mirror / Atom feed
From: Benjamin Cronce <bcronce@gmail.com>
To: George Amanakis <g_amanakis@yahoo.com>
Cc: cake@lists.bufferbloat.net
Subject: Re: [Cake] WAN ingress rate with concurrent downloads
Date: Sat, 10 Dec 2016 13:33:35 -0600	[thread overview]
Message-ID: <CAJ_ENFFF_Z=oOnWRKQtmMcaZtCmZcOT3FE6idnhCcJ+8=u5=+A@mail.gmail.com> (raw)
In-Reply-To: <1462124589.1190968.1481310707746@mail.yahoo.com>

[-- Attachment #1: Type: text/plain, Size: 1106 bytes --]

It sounds like the sender has a buffer bloat issue. I see the same thing
with bittorrent and tcp. If I rate limit my 150mb connection to 100mb, I
will see about 30-40mb of retransmissions, for a total of ~135mb of
ingress. Even though I limit only to 100mb.

If I trace route the offending senders, I always see a 3000ms+ ping 1-2
hops before I reach them.

Tcp treats the buffer bloat as packet loss, flooding the receiver.

On Dec 9, 2016 1:20 PM, "George Amanakis" <g_amanakis@yahoo.com> wrote:

> Dear All,
>
> regarding the issue about WAN ingress rate with many concurrent TCP
> downloads, it seems that all the excess ingress rate on the WAN interface
> are TCP retransmission packets (Wireshark). After the latest commit 78ff814
> in cobalt, ping times while using TCP BitTorrent improved from ~1000ms to
> ~300ms (concurrent connections 55). If I reduce the concurrent connections
> to 30 I get ping times lower than 70ms.
>
> Best regards,
> George
> _______________________________________________
> Cake mailing list
> Cake@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cake
>

[-- Attachment #2: Type: text/html, Size: 1653 bytes --]

  reply	other threads:[~2016-12-10 19:33 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1462124589.1190968.1481310707746.ref@mail.yahoo.com>
2016-12-09 19:11 ` George Amanakis
2016-12-10 19:33   ` Benjamin Cronce [this message]
     [not found] <mailman.1.1477670401.20617.cake@lists.bufferbloat.net>
2016-10-30 16:43 ` Georgios Amanakis
2016-10-30 19:27   ` Sebastian Moeller
     [not found] <mailman.5.1477497601.28053.cake@lists.bufferbloat.net>
2016-10-27  0:10 ` G. Amanakis
2016-10-27  0:16   ` G. Amanakis
2016-10-27  6:41     ` Sebastian Moeller
2016-10-27 16:42       ` Noah Causin
2016-10-28  1:34     ` Georgios Amanakis
2016-10-28  7:45       ` Sebastian Moeller

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/cake.lists.bufferbloat.net/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAJ_ENFFF_Z=oOnWRKQtmMcaZtCmZcOT3FE6idnhCcJ+8=u5=+A@mail.gmail.com' \
    --to=bcronce@gmail.com \
    --cc=cake@lists.bufferbloat.net \
    --cc=g_amanakis@yahoo.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