netdev - Re: ingress policying for realtime protocol

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAHA+R7N8wFKPzOR0EshWzcaTUd+ChMW0zb0=1go4oV_gxoAyiA@mail.gmail.com>
Date:	Wed, 20 May 2015 16:46:54 -0700
From:	Cong Wang <cwang@...pensource.com>
To:	Uwe Kleine-König 
	<u.kleine-koenig@...gutronix.de>
Cc:	netdev <netdev@...r.kernel.org>,
	Jamal Hadi Salim <jhs@...atatu.com>
Subject: Re: ingress policying for realtime protocol

(Cc'ing Jamal)

On Tue, May 19, 2015 at 2:11 PM, Uwe Kleine-König
<u.kleine-koenig@...gutronix.de> wrote:
> Hello,
>
> I hope this is the/a right list to ask this question. If not, I'd be
> happy to learn the right alternative.
>
> For a system that implements the Media Redundancy Protocol (MRP,
> http://en.wikipedia.org/wiki/Media_Redundancy_Protocol) I'm faced with
> the task to assert that the corresponding ethernet packets are handled
> "in time". Currently this fails already if I do
>
>         scp somebigfile mymrpsystem:
>
> The first idea was to use ingress policying to limit all packets that
> are not related to MRP to a certain rate.
>
> The problem in practise however is that the rate that has to be used
> above is very low (100kbit) so we're looking for alternatives. I wonder
> what's missing to get ingress policying more versatile to allow for
> example a prio qdisc or if there is some conceptual problem that I don't
> see. I'm not sure this would solve our problem, but it's definitely
> something I want to play with before trying/resorting to hackish
> solutions (like hacking the eth-driver to pipe MRP packets directly to
> the network userspace application instead of through the network stack).
>

There is very little to do on ingress side since there is no queue at all,
not to mention priority, you could try ifb to see if it fits your need.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html