From: Claudio Greco <cla.greco@fastwebnet.it>
To: lartc@vger.kernel.org
Subject: Re: [LARTC] how hierarchical is HTB?
Date: Wed, 06 Jun 2007 14:58:14 +0000 [thread overview]
Message-ID: <4666CB86.2070304@fastwebnet.it> (raw)
In-Reply-To: <20070606111619.26ef7781@pulsar.inexo.com.br>
> root class 1: (rate\x100, ceil\x100)
> 1: children classes 1:10 (30,100) and 1:20 (70,100)
> 1:10 children classes 1:100 (10,100) and 1:101 (20,100)
> 1:20 children classes 1:200 (30,100) and 1:201 (70,100)
>
> I managed to have the root rate equals to the sum of its children.
>
>
Well, it is still true that total assured rate for classes 1:200 and
1:201 is greater than assured rate for class 1:20. Still, I don't think
this is a big deal.
> But how must the rates of the leaves be signed?
>
What do you mean with 'signed'?
> And how the bandwidth of these leaves will be distributed when
> borrowing/lending is necessary?
>
>
As far as I know, when a leaf is 'yellow', i.e. its rate is greater than
its assured rate and lesser than its ceil rate, it can borrow from its
parent providing there's a yellow-path to the root and the root is green
(root can't be yellow, only green or red).
If there's more than one child borrowing from the same class, they're
served according to their priority (argument prio in *tc class add*).
If there's more than one child having the same priority, then they're
served in DRR order (Deficit Round Robin).
You can tune DRR behaviour with arguments r2q in *tc qdisc add* and
quantum in *tc class add*.
> classs 1:10 will/may lend/borrow from class 1:20. I know that.
>
No it can not. A class can only borrow from its parent, never from its
siblings.
> But how about 1:1XX and classes 1:2XX? will the borrow/lend from each
> others?
>
>
ibidem.
> Any docs about this?
>
>
You may see:
http://luxik.cdi.cz/~devik/qos/htb/manual/userg.htm
http://luxik.cdi.cz/~devik/qos/htb/manual/theory.htm
_______________________________________________
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
next prev parent reply other threads:[~2007-06-06 14:58 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-06-06 14:16 [LARTC] how hierarchical is HTB? Ethy H. Brito
2007-06-06 14:58 ` Claudio Greco [this message]
2007-06-06 16:49 ` Ethy H. Brito
2007-06-06 20:44 ` Flechsenhaar, Jon J
2007-06-09 7:24 ` Marco Aurelio
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=4666CB86.2070304@fastwebnet.it \
--to=cla.greco@fastwebnet.it \
--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.