From: Sebastian Moeller <moeller0@gmx.de>
To: Jonathan Morton <chromatix99@gmail.com>,
Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>,
Jonathan Morton <chromatix99@gmail.com>,
Kevin Darbyshire-Bryant <kevin@darbyshire-bryant.me.uk>
Cc: bloat <bloat@lists.bufferbloat.net>
Subject: Re: [Bloat] ADSL, ATM drivers, bloat, education & confusion
Date: Sat, 06 Jun 2015 17:04:48 +0200 [thread overview]
Message-ID: <AC3872B8-209C-4416-9A21-94A90A911A88@gmx.de> (raw)
In-Reply-To: <CAJq5cE1LWz_LMb8rGLMuU5doQZw9uqxWgL=BMi0m=r4DAe3TcQ@mail.gmail.com>
Hi Jonathan, hi Kevin, hi list,
On June 6, 2015 3:53:43 PM GMT+02:00, Jonathan Morton <chromatix99@gmail.com> wrote:
>Ideally, one would read the raw sync rate out of the modem (there has
>to be
>a way to do that, since it is able to report it in the web interface),
>and
>use that to set cake's bandwidth parameter.
Unfortunately ISPs more and more implement a throttle/policed? at the BRAS level which in essence supersedes the link rate, but does not seem to be universally reported. If ISPs would actually use this in a way that keeps utilization high but latency increase low all would be dandy, just alas they might be trying but at least not all of them succeed. So people should start out at the modems link rates but should be prepared to reduce the shaped rates until latency under load seems bounded to acceptable levels...
Best Regards
Sebastian
You could update it
>periodically using "tc qdisc change ...".
>
>You will need to set the overhead calculation correctly, probably using
>"pppoa-vcmux" or similar, depending on precisely what encapsulation
>your
>ISP uses. If you're not confident about that, specify "conservative" to
>begin with. Once you have that, you can experiment to see how much you
>need
>to reduce the raw sync rate to achieve full bottleneck control -
>hopefully
>not much, at most 1%.
>
>The same approach would also work for inbound if necessary, except that
>you
>would need to reduce the raw sync rate rather more.
>
>- Jonathan Morton
>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>Bloat mailing list
>Bloat@lists.bufferbloat.net
>https://lists.bufferbloat.net/listinfo/bloat
--
Sent from my Android device with K-9 Mail. Please excuse my brevity.
next prev parent reply other threads:[~2015-06-06 15:04 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-06-06 13:38 Kevin Darbyshire-Bryant
2015-06-06 13:53 ` Jonathan Morton
2015-06-06 14:30 ` Sebastian Moeller
2015-06-11 11:24 ` Jesper Dangaard Brouer
2015-06-11 11:57 ` Sebastian Moeller
2015-06-11 14:14 ` Tristan Seligmann
2015-06-11 21:03 ` Sebastian Moeller
2015-06-12 6:45 ` Jesper Dangaard Brouer
2015-06-06 15:04 ` Sebastian Moeller [this message]
2015-06-06 14:29 ` Sebastian Moeller
2015-06-06 15:46 ` Kevin Darbyshire-Bryant
2015-06-06 15:50 ` Dave Taht
2015-06-06 18:14 ` Sebastian Moeller
2015-06-07 13:29 ` Kevin Darbyshire-Bryant
2015-06-11 11:16 ` Jesper Dangaard Brouer
2015-06-11 11:51 ` 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/bloat.lists.bufferbloat.net/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=AC3872B8-209C-4416-9A21-94A90A911A88@gmx.de \
--to=moeller0@gmx.de \
--cc=bloat@lists.bufferbloat.net \
--cc=chromatix99@gmail.com \
--cc=kevin@darbyshire-bryant.me.uk \
/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