CoDel AQM discussions
 help / color / mirror / Atom feed
From: Dave Taht <dave.taht@gmail.com>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Eric Dumazet <edumazet@google.com>,
	codel@lists.bufferbloat.net,
	Dave Taht <dave.taht@bufferbloat.net>
Subject: Re: [Codel] [PATCH 1/2] codel: Controlled Delay AQM
Date: Mon, 7 May 2012 10:52:29 -0700	[thread overview]
Message-ID: <CAA93jw60g4Jt6RkvS9cEL1x+_7bRSQ547cJ3ObSx-2tBCjXw=A@mail.gmail.com> (raw)
In-Reply-To: <1336412079.3752.2333.camel@edumazet-glaptop>

On Mon, May 7, 2012 at 10:34 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Mon, 2012-05-07 at 10:27 -0700, Rick Jones wrote:
>
>>
>> I don't wish to rathole, and someone may want to declare Bikeshed! but
>> the target audience of the paper is a bit different from the Linux
>> system administrator base yes?  If a parameter name enhancement makes
>> something more clear to an administrator is it really a bad thing to
>> have a name that is not a strict duplicate of the paper's?
>>
>> rick
>
> I dont know, we have no precedence in linux for 'target' and 'interval'
>
> So picking the codel words seems ok

I like short names from a readability perspective. I'm not big on underscores,
they are hard to see. (I'm getting old)

I note that codel is intended to run in most cases without any parameters.

limit: used elsewhere in linux for a packet limit. A packet limit is there for
practical reasons, shrinking the queue has overhead. However, codel will
run just fine with infinite queue (yea!), and I suspect the default
1000 is too small
on 10GigE. If there was a way to autotune it based on the underlying
device, that would be good.

target: At the moment this is a knob because 5ms is an eternity at 10GigE
interval: Same reason

I'm not big on changing these from the paper, nor on _s, nor on length.
However, I'm open to further discussion.

maxpacket: really confusing usages, so I went with minbytes which is explicit

Any objections?

(min was overloaded with red's usage, so I felt that would be confusing, too)

ecn:

ecn was not in the paper (and it's working well) and I'd like to live
in a world where ecn was the default. I think. Others disagree.

so perhaps ecn and noecn? And ecn be the default (after some serious testing?)

Parameterless is a good goal.

>
>
>
> _______________________________________________
> Codel mailing list
> Codel@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/codel



-- 
Dave Täht
SKYPE: davetaht
US Tel: 1-239-829-5608
http://www.bufferbloat.net

  reply	other threads:[~2012-05-07 17:52 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-05-05 19:32 Dave Taht
2012-05-05 19:32 ` [Codel] [PATCH 2/2] codel: RED is dead Dave Taht
2012-05-05 19:47 ` [Codel] [PATCH 1/2] codel: Controlled Delay AQM Dave Taht
2012-05-05 20:05   ` Eric Dumazet
2012-05-07 16:43 ` Rick Jones
2012-05-07 17:10   ` Eric Dumazet
2012-05-07 17:18     ` Dave Taht
2012-05-07 17:27     ` Rick Jones
2012-05-07 17:34       ` Eric Dumazet
2012-05-07 17:52         ` Dave Taht [this message]
2012-05-07 18:01           ` Eric Dumazet
2012-05-07 18:09             ` Eric Dumazet
2012-05-07 18:15               ` Eric Dumazet
2012-05-07 18:32           ` Jim Gettys
2012-05-07 18:44             ` Dave Taht
2012-05-07 18:50               ` Jim Gettys
2012-05-07 19:16                 ` Eric Dumazet
2012-05-07 19:36                   ` Jim Gettys
2012-05-07 20:03                     ` Eric Dumazet
2012-05-07 20:55                       ` dave taht
2012-05-08  1:15                       ` Jim Gettys
2012-05-08  1:20                         ` Andrew McGregor
2012-05-08  1:45                           ` Jim Gettys
2012-05-08  2:16                             ` Rick Jones
2012-05-08  3:14                               ` Jim Gettys
2012-05-08  3:32                                 ` Rick Jones
2012-05-08  3:40                                   ` Kathleen Nichols
2012-05-08  4:14                                   ` Jim Gettys
2012-05-08  5:10                                     ` Eric Dumazet
2012-05-08 10:25                                       ` Jim Gettys
2012-05-08 13:57                                         ` Kathleen Nichols
2012-05-08 14:04                                           ` Jim Gettys

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='CAA93jw60g4Jt6RkvS9cEL1x+_7bRSQ547cJ3ObSx-2tBCjXw=A@mail.gmail.com' \
    --to=dave.taht@gmail.com \
    --cc=codel@lists.bufferbloat.net \
    --cc=dave.taht@bufferbloat.net \
    --cc=edumazet@google.com \
    --cc=eric.dumazet@gmail.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