From: warren ponder <wponder11@gmail.com>
To: Andrew Somerville <aesomerville@gmail.com>
Cc: Sebastian Moeller <moeller0@gmx.de>,
Starlink-Bufferbloat <starlink@lists.bufferbloat.net>
Subject: Re: [Starlink] CAKE-Autorate
Date: Fri, 2 Dec 2022 06:19:45 -0700 [thread overview]
Message-ID: <CACnq21cBhDQYVjkCdAn1E3Qm0R06tDZDM1oD9sEU3F=k5oOgTQ@mail.gmail.com> (raw)
In-Reply-To: <CAMUVRMaALXFC0VpivGu2c+Cn1zSNcb=oAik2agrWvzv+LR1vqw@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 4018 bytes --]
Thanks Andrew, I sent @moeller some logs but the data set was small. I
should have a better one this AM. I can send it here or Openwrt ( probably
better) I do not want to clog up this list.
I'll check out the link and see if I have the reccomendations implemented.
Warren
On Fri, Dec 2, 2022, 6:09 AM Andrew Somerville <aesomerville@gmail.com>
wrote:
> Dear Warren,
>
> As Sebastian suggested if you can send across a log file having run some
> speed tests (ideally with longer rather than shorter duration) then with
> his fantastic plotting utility we should be able to get an idea of what is
> happening. There really is not much in the way of documentation concerning
> cake-autorate. I have put almost all my effort into developing it with the
> help of Sebastian.
>
> cake-autorate does not autoconfigure itself so it does require quite a bit
> of tuning - especially for Starlink, which is a super interesting case
> because it offers huge bandwidth but latency is a challenge to keep under
> control and we even tried to implement some compensation for satellite
> switching, which hasn't been tested as much as I would like and possibly
> can be improved further.
>
> It might be time to add some more documentation. But the thread on OpenWrt
> is a great place to get feedback from both of us and other users. Actually
> I just gave some suggested parameters for use with Starlink to an OpenWrt
> user here:
>
> https://forum.openwrt.org/t/cake-w-adaptive-bandwidth/135379/1221?u=lynx
>
> Best wishes and keep me CC'ed,
>
> Andrew
>
> On Thu, 1 Dec 2022 at 20:52, warren ponder <wponder11@gmail.com> wrote:
>
>> Thanks @moeller! I am still trying to figure out Octave. It's not
>> running on my router so I have been copying zips off to a Mac and
>> trying to plot them there. I think I am not hitting the min
>> requirement of 6.4 though. Seems the plots are key here so I'll put
>> some effort into focusing on that.
>>
>> WP
>>
>> On Thu, Dec 1, 2022 at 1:40 PM Sebastian Moeller <moeller0@gmx.de> wrote:
>> >
>> > Hi Warren,
>> >
>> >
>> > > On Dec 1, 2022, at 19:41, warren ponder via Starlink <
>> starlink@lists.bufferbloat.net> wrote:
>> > >
>> > > I have been working with the 3rd party router team of the router I
>> > > have been using behind Starlink to add CAKE to their router. I have an
>> > > early access loaded and have modified it to add CAKE-Autorate. I have
>> > > it running but not entirely sure I have it all running correctly or
>> > > how to quantify if its making a difference.
>> > >
>> > > Beside github for CAKE-Autorate are there any other sources of info to
>> > > read up on about ensuring I have it going right and possibly quantify
>> > > its making a difference?
>> >
>> > Most discussion happens in the following OpenWrt thread:
>> > https://forum.openwrt.org/t/cake-w-adaptive-bandwidth/135379/1202
>> > or you can open an issue in the github. About making a difference we
>> added some potting code which can turn the log files into partially
>> interpretable plots.
>> >
>> > I use (192.168.42.1 is my OpenWrt/TurrisOS router running
>> cake-autorate):
>> > ssh root@192.168.42.1 'kill -USR1 $( cat
>> /var/run/cake-autorate/maintain_log_file_pid )' && sleep 30 && scp
>> root@192.168.42.1:/var/log/cake-autorate_*.log.gz ./SCRATCH/ && ssh
>> root@192.168.42.1 'rm /var/log/cake-autorate_*.log.gz'
>> >
>> > followed by:
>> > octave -qf --eval
>> 'fn_parse_autorate_log("./SCRATCH/cake-autorate_2022-12-01_20_21_16.log.gz",
>> [], [])'
>> >
>> > this requires octave >= 6.4... once you have that post your plots to
>> the forum thread and we can take it from there.
>> >
>> >
>> >
>> > Regards
>> > Sebastian
>> >
>> > P.S.: I cc'd Andrew cake-autorate's principal developer, he might have
>> more ideas...
>> >
>> >
>> >
>> > >
>> > > thanks
>> > >
>> > > WP
>> > > _______________________________________________
>> > > Starlink mailing list
>> > > Starlink@lists.bufferbloat.net
>> > > https://lists.bufferbloat.net/listinfo/starlink
>> >
>>
>
[-- Attachment #2: Type: text/html, Size: 5910 bytes --]
prev parent reply other threads:[~2022-12-02 13:19 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-12-01 18:41 warren ponder
2022-12-01 20:40 ` Sebastian Moeller
2022-12-01 20:52 ` warren ponder
2022-12-01 21:25 ` Sebastian Moeller
2022-12-02 13:08 ` Andrew Somerville
2022-12-02 13:19 ` warren ponder [this message]
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/starlink.lists.bufferbloat.net/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CACnq21cBhDQYVjkCdAn1E3Qm0R06tDZDM1oD9sEU3F=k5oOgTQ@mail.gmail.com' \
--to=wponder11@gmail.com \
--cc=aesomerville@gmail.com \
--cc=moeller0@gmx.de \
--cc=starlink@lists.bufferbloat.net \
/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