All of lore.kernel.org
 help / color / mirror / Atom feed
From: raptor <raptor@tvskat.net>
To: lartc@vger.kernel.org
Subject: [LARTC] shaping combined traffic ?
Date: Thu, 20 Feb 2003 09:07:50 +0000	[thread overview]
Message-ID: <marc-lartc-104573207010209@msgid-missing> (raw)

I have the following question.

Is there a way to shape fairly trought one big channel two others, what I have in mind.


users <-------- [eth2]<===[eth0]<------ Internet (1Mbit/s)
	 			      |==[eth1]<------ Peering (1Mbit/s)

Now the problem, as u see there is two internet downstrams with compound speed of 2Mbit/s, the poblem is that on the eth2-egrees there should be only one class(channel), sum of the others. Say for example all users have 100kb/s (so that we can calculate easly). Now I want to make a compound channels so that when internet-channels are free users can get unused speed (ceil), (on the other hand rate calculation get very complicated if imposible 'cause it may exceed a wrapper-channel, so that some mechanism has to be thought to do this). Here is the pseudo code :


dev eth2 {
	htb(rate 2Mbit/s) {
		class(rate 100kbps, ceil 150kbps);
		class(rate 100kbps, ceil 150kbps);
		class(rate 100kbps, ceil 150kbps);
		.....
		class(rate 100kbps, ceil 150kbps);
	}
}

oops to metion that one of channels is national peering and the other is internet.
So as u see the compound speed can easly exceed one of the channels w/o ever touching the second channel.. I mean that if for example all users start to dload only from the peering channel, they will exceed easly the 1Mbit/s capacity,  'cause wrapper channel is 2MBit/s
On the other hand if I don't put this 2mbit-wrapper I can't use ceil-functionality.
Also I can't create 2 wrapper-channels 1Mbit/s each, 'cause users has to have 100kb/s compound speed not 100kb/s peering and 100kb/s internet,  neither 50kb/s-peering and 50kb/s-internet..
The 100kb/s is divided depending on their current usage, so it is not predefined..

So the questions are - how to have total speed X divided by all channels dynamicly and in the same time be able to use ceil-functionality.

Thanx alot in advance for your ideas

raptor
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

                 reply	other threads:[~2003-02-20  9:07 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=marc-lartc-104573207010209@msgid-missing \
    --to=raptor@tvskat.net \
    --cc=lartc@vger.kernel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.