From: Sebastian Moeller <moeller0@gmx.de>
To: Stephen Hemminger <stephen@networkplumber.org>
Cc: Jan Rovner via Cake <cake@lists.bufferbloat.net>
Subject: Re: [Cake] Cake - flow (or host) weight request - how to implement
Date: Wed, 27 Sep 2023 22:44:51 +0200 [thread overview]
Message-ID: <219B7A03-E587-43AF-8A98-B7872B3B525B@gmx.de> (raw)
In-Reply-To: <20230927132732.6585172a@hermes.local>
Hi Stephen,
> On Sep 27, 2023, at 22:27, Stephen Hemminger via Cake <cake@lists.bufferbloat.net> wrote:
>
> On Wed, 27 Sep 2023 17:48:06 +0000
> Jan Rovner via Cake <cake@lists.bufferbloat.net> wrote:
>
>> Hello,
>>
>> I am trying to experiment with sch_cake.c, my goal is to ad-hoc prioritize / deprioritize traffic served to specified a host(s). The prioritization would be controlled by an external system and it should handle well the situation where a host has active flow(s).
>
> Is it possible to use one or more existing BPF hooks to change DCSP bits for that paths you wish to adjust.
> The point is that changes to upstream base code are hard to maintain, and that is what BPF initial hooks
> were meant to provide.
IIUC the OP wants not only to steer packets into a specific tier but really wants to move a hash-bins complete state between tiers (he talks about cobalt state). My (limited) understanding is that this would need changes to cake's kernel code, either by always keeping shadow cobalt state per hash-bin in all tiers or by implementing a state copy procedure...
Now, I am not sure how well moving flows around priority tiers is going to work on a timescale that preserving the cobalt state seems important and how to avoid re-ordering if a a flow is increased in priority while it still has queued packets in the lower priority tier (maybe the solution is moving both cobalt state and queued packets between tiers).
Regards
Sebastian
> _______________________________________________
> Cake mailing list
> Cake@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cake
prev parent reply other threads:[~2023-09-27 20:44 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-27 17:48 Jan Rovner
2023-09-27 20:27 ` Stephen Hemminger
2023-09-27 20:44 ` Sebastian Moeller [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/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=219B7A03-E587-43AF-8A98-B7872B3B525B@gmx.de \
--to=moeller0@gmx.de \
--cc=cake@lists.bufferbloat.net \
--cc=stephen@networkplumber.org \
/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