[3437] in linux-net channel archive
Re: shaper or whatever
daemon@ATHENA.MIT.EDU (Miquel van Smoorenburg)
Mon Jun 24 01:00:20 1996
To: submit-linux-dev-net@ratatosk.yggdrasil.com
From: Miquel van Smoorenburg <miquels@cistron.nl>
Date: Mon, 24 Jun 1996 01:43:01 +0200 (MET DST)
In article <m0uXs2w-0005FaC@lightning.swansea.linux.org.uk>,
Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
>
>It may have been me (and Im not an academic so nyaahh ;))
Come on, it's getting to look to look like Linus is the only
one with a diploma in the Linux world ;)
>> >I understand from Alan Cox that someone is already writing this for linux.
>> >We shall see when it arrives.
>> It is designed as above? The concept is feasible....but the above is
>> someone leaky in concept.
>
>The stuff Im doing is based directly on Van's token bucket filter that does
>the multicast traffic shaping, only thinned down and faster (ok its faster
>and crashes your PC just at the moment.. 8)).
Ah. Are you basing it on the firewall code? I want to start to implement
something based on the firewall code. It adds a bandwidth argument to
ipfwadm and sets up a bandwidth struct per source/destination specified.
This is also recorded in the skbuff. Then in the drivers io routines
this is checked per skbuff and the bandwidth is limited. This is done
by calculating how long it would take to transmit the packet with the
virtual bandwidth available and only allowing a new packet to be
transmitted after that period has passed (using timers etc). If the
backlog grows too big packets will be dropped. This behaves just like
a normal serial link, but allowing several in parallel.
I have played with this before as you can see when ftp'ing to
ftp.cistron.nl (3 kb bandwidth shared between all anonymous
ftp logins ;))
Does this concept sound sane? Is this like the thing you are implementing?
Mike.
--
+ Miquel van Smoorenburg + Cistron Internet Services + Living is a |
| miquels@cistron.nl (SP6) | Independent Dutch ISP | horizontal |
+ miquels@drinkel.ow.org + http://www.cistron.nl/ + fall +
